blitzscaling: the lightning-fast path to building massively valuable companies
TRANSCRIPT
Copyright©2018byReidHoffmanandChrisYehAllrightsreserved.PublishedintheUnitedStatesbyCurrency,animprintoftheCrownPublishingGroup,adivisionofPenguinRandomHouseLLC,NewYork.currencybooks.com
CURRENCYanditscolophonaretrademarksofPenguinRandomHouseLLC.
LibraryofCongressCataloging-in-PublicationDataNames:Hoffman,Reid,author.|Yeh,Chris,author.Title:Blitzscaling:thelightning-fastpathtocreatingmassivelyvaluablebusinesses/byReidHoffmanandChrisYeh.Description:Firstedition.|NewYork:Currency,[2018]|Includesbibliographicalreferencesandindex.Identifiers:LCCN2017058413|ISBN9781524761417Subjects:LCSH:Newbusinessenterprises.|Smallbusiness—Growth.|Entrepreneurship.Classification:LCCHD62.5.H6242018|DDC658.4/063—dc23LCrecordavailableathttps://lccn.loc.gov/2017058413
ISBN 9781524761417EbookISBN 9781524761424
CoverdesignbyJoshSmith
v5.3.2
ep
Contents
CoverTitlePageCopyright
ForewordbyBillGates
Introduction
PartI:WhatIsBlitzscaling?SoftwareIsEating(andSaving)theWorldTheTypesofScalingTheThreeBasicsofBlitzscalingTheFiveStagesofBlitzscalingTheThreeKeyTechniquesofBlitzscaling
PartII:BusinessModelInnovationDesigningtoMaximizeGrowth:TheFourGrowthFactorsDesigningtoMaximizeGrowth:TheTwoGrowthLimitersProvenBusinessModelPatternsTheUnderlyingPrinciplesofBusinessModelInnovationAnalyzingaFewBillion-DollarBusinessModels
PartIII:StrategyInnovationWhenShouldIStarttoBlitzscale?WhenShouldIStopBlitzscaling?CanIChooseNottoBlitzscale?BlitzscalingIsIterativeHowBlitzscalingStrategyChangesinEachStage
HowtheRoleoftheFounderChangesinEachStage
PartIV:ManagementInnovationEightKeyTransitionsNineCounterintuitiveRulesofBlitzscalingTheNever-endingNeedforChange
PartV:TheBroaderLandscapeofBlitzscalingBlitzscalingBeyondHighTechBlitzscalingWithinaLargerOrganizationBlitzscalingBeyondBusinessBlitzscalinginGreaterSiliconValleyOtherBlitzscalingRegionstoWatchChina:TheLandofBlitzscalingDefendingAgainstBlitzscaling
PartVI:ResponsibleBlitzscalingBlitzscalinginSocietyFrameworkforResponsibleBlitzscalingTheResponseSpectrumBalancingResponsibilityandVelocityastheOrganizationGrows
ConclusionAcknowledgmentsAppendixA:DisclosuresAppendixB:TheBlitzscalersAppendixC:CS183CEssays
FOREWORDBYBILLGATES
I’ve known Reid Hoffman for years. Our friendship started on my visits toSiliconValley tomeetwithGreylockPartners, the venture capital firmwhereReidisapartner,soIcouldlearnaboutthecompaniestheywereinvestingin.Iwas always impressed by his sharpmind and brilliant business sense. Reid isfamousforhostinglongdinnerswheretheconversationrunslateintothenight,andwe’vespentmanymealsbreakingdownthetechnologyindustry,analyzingthe promise of artificial intelligence, and more. When Microsoft CEO SatyaNadellastartedtalkingaboutacquiringLinkedIn,Iknewitwouldbeanamazingfit.OfallthethingsI’vediscussedwithReid,themostthought-provokingmight
beblitzscaling.It isanideathatappliestomanydifferentindustries,asheandChris explain in the last section of this book. But prioritizing speed overefficiency—evenin thefaceofuncertainty—isespecially importantwhenyourbusinessmodeldependsonhaving lotsofmembersandgetting feedback fromthem. If you get in early and start getting that feedback and your competitorsdon’t, then you’re on the path to success. In any business where scale reallymatters,gettinginearlyanddoingitfastcanmakethedifference.This is especially true for two-sided businessmodels, where you have two
user groups that create positive network effects for each other. For example,LinkedInwantstoattractpeoplewhoarelookingforworkaswellasemployerswhowanttohirethem.Airbnbwantsguestslookingforaplacetostayaswellashostswithspacetorent.Uberwantstoattractdriversaswellasriders.And a software company with an operating system to sell wants app
developersaswellasendusers.Microsoftdefinitelywentthroughablitzscalingphase(althoughwedidn’tcallitthatatthetime).Wegotonthelearningcurveearly and were able to build a reputation as a serious company. We had anextremecultureofworkinghardandgettingthingsdonefast.The ideas behind blitzscaling aren’t just for startups and scale-ups. They’re
importantforbig,establishedcompaniestoo.Thewindowforactioncanbetinyand it can close quickly. Even a few months of hesitation can mean thedifferencebetweenleadingandchasing.ReidandChris’sideasaremorepracticalthanever,becauseitisnowpossible
togetbigfastinawaythatsimplywasn’tfeasibleafewdecadesago.Thereisarichecosystemofserviceprovidersandoutsourcingcompaniestosupportrapidgrowth.Many companies have gone through their own big growth spurts, sothere are lots of examples to learn from. User feedback comes in a constantstreamofdata.Productcycleshavedroppedfromyearlytoweeklyordaily.Andgood reviews can spread in an instant online, so a strong product can quicklyattractabigaudience.Inotherwords, thecasestudiesyou’reabouttoexploreandthetoolsyou’re
about to gain have never been more relevant. This is an ideal moment to bereadingthisbook.I’mgladReidandChrisaresharingtheirinsights.
Introduction
2011:SANFRANCISCO,AIRBNBHEADQUARTERS
“They’reprobablygoingtokillyou.”Theyearwas2011,and in theofficesofAirbnb, thenascrappy little forty-
person start-up, its cofounder and CEOBrian Chesky had just received someverybadnews.Brian pondered the implications of the ominous prediction he’d just heard
fromAndrewMason,thecofounderandCEOofGroupon.Hedidn’tlikeit.Brian and his cofounders, JoeGebbia andNathanBlecharcyzk, had already
fought their way through plenty of obstacles to build Airbnb, a website thatmakes iteasy forpeople to rentout their roomsorhomesfor thenight. In thebeginning, every investor the founders approached had turned them down or,worse, ignored them. The companywas on the upswing now, but the painfulearlydayswerestill fresh in theirminds,andtheyweren’t lookingforanotherbattle.
WhentheAirbnbfoundersfirstmet,PaulGraham,thehighlyregardedfounderofthestart-upacceleratorYCombinator(YC),toldthemflatoutthattheirideawas terrible. “People are actually doing this?!” he incredulously asked.WhenBrian told him yes, peoplewere, in fact, renting out their living spaces for anight,Graham’sresponsewas“What’swrongwiththem?”Still, Graham had accepted the Airbnb guys into the three-month-long YC
program.NotbecausehewasinspiredbytheirAirbnbbusiness,butbecausehewas impressedby thehustleof thefounders.He loved the(nowfamous)storyabouthowCheskyandhiscofoundersmanagedtopaythebillswhiletryingtogetAirbnbofftheground.Itwas2008,aUSpresidentialelectionyear,sothey
created and sold special-edition cereals called “Obama O’s” and “Cap’nMcCains”—asugaryparodyof(ortributeto,dependingonhowyoulookatit)that year’s candidates Barack Obama and John McCain. The creativity andpersistencedisplayedbytheAirbnbfoundersas“cerealentrepreneurs”gottheminthedooratYC;onceintheprogram,theyrefinedtheirbusinessandwereableto persuade two leading venture capital firms, Sequoia Capital and GreylockPartners(whereIamageneralpartner),toinvest.Now, nearly four years later, it seemed like all the hard work was finally
startingtopayoff.Havingcelebrateditsmillionthbooking,Airbnbhadplentyofworkingcapital,anditwasclearthattheconceptwasvaluable.But when you’re successful, you attract competition. And sometimes that
competitionrepresentsadeadlythreat.In Airbnb’s case, that threat was three brothers from Cologne, Germany:
Oliver, Marc, and Alexander Samwer. They had become billionaires byanalyzingsuccessfulUScompanies,rapidlycreatingcopycatsinEurope,and,inmany cases, selling those “cloned” companies to their original Americaninspirations.Inothercases, theSamwersactuallyheldontoandbuiltout theirclones;Zalando,the“ZapposofEurope,”hadovertenthousandemployeesandwasworthmorethan$10billionin2017.TheirfirstsuccesswasAlando,aneBayknockoffthattheywereabletosellto
eBay for $43 million, just one hundred days after launching it. The SamwerbrotherstheninvestedintheGermanversionsofYouTube(MyVideo),Twitter(Frazr), and Facebook (StudiVZ) before founding their own start-up studio,RocketInternet.In early 2011, Brian and his team started noticing that Airbnb users were
beingspammedbyanewcompanynamedWimdu.Wimduhadapparentlyjustreceived $90million—the largest investment in a European start-up to date—fromnoneotherthanRocketInternetandKinnevik,amajorSwedishinvestmentcompanythathadpartneredwiththeSamwerbrothers.Theproblem?Wimdu’sbusinessmodelandwebsitelookedlikeaknockoffof
Airbnb’s.Wimdu was founded in March 2011, and, within weeks, the Berlin-based
company had hired a staggering four hundred employees and opened twentyoffices acrossEurope.Meanwhile, the original, butmuch smaller,Airbnb hadraised only $7million, had just forty employees, and operated out of a singleofficeinSanFrancisco.Asafirst-timeCEO,Brianwasn’tevensurewhatwas
involvedinopeningasecondoffice,letalonedozensmoreonanothercontinent.Brian also knew that if Wimdu was able to capture and dominate the
European market, Airbnb might not survive. “If you’re a travel site and youdon’t cover Europe, you’re dead,” he told us in 2015, when he visited theTechnology-EnabledBlitzscalingclasswetaughtatStanfordUniversity.The Samwer brothers had named their price: Airbnb could haveWimdu in
exchangefora25-percentstakeinAirbnb.NowBrianfacedadifficultdecision,withpainfulconsequencesregardlessofwhatoptionhechose.In response,Brian turned tooneofhis favoritedecision-making techniques:
reaching out to the world’s leading experts. His first call was to the CEO ofGrouponatthetime,AndrewMason.Theleadingdailydealscompanyhadhadasimilarexperiencethepreviousyear:InDecember2009,theSamwerbrothershadlaunchedCityDeal,theirGrouponlookalike.Sixmonthslater,Grouponpaidanine-figureprice, roughly10percentof itsvaluationat thatpoint, toacquirethiscompetitor.HerewasthequestionweighingheavilyonBrianandhisteam:ShouldAirbnb
follow Groupon’s strategy and just buy the knockoff company? Brian’s gutinstinct was to say no. IntegratingWimdu’s finance-centric andmetric-driventeamcouldharmAirbnb’sdesign-drivenculture.Hewasalsoreluctanttorewardwhat he saw as a legal extortion racket rather than a sincere attempt to createvalueinthemarket.YetBrianfelthehadanobligationtoconsidertheoffer.Masonhadtoldhim
thatdespitethemanyproblemstheCityDealacquisitionhadbrought,ithadalsoaccelerated Groupon’s progress into the European market, which ended upaccountingfornearly30percentofitsglobalsales.Itcouldeasilybearguedthatgiving up 10 percent of Groupon for CityDeal was actually a good deal. Butperhaps emboldened by their successful CityDeal gambit, the Samwers wereaskingforafarlargershareofAirbnb—afull25percent.On the other hand, Airbnb could reject the offer and instead take on the
aggressiveSamwerbrothersinahead-to-headcompetition.ButWimduhadthehome-turf advantage, not to mention ten times the number of employees andmore than ten times the amount of invested capital. Competing against themwouldbeonehellofanuphillbattle.Tiredofthefund-raisinggrind,especiallyitsemotionaltoll,Brianwondered
whetherhehaditinhimtotakeonthisnewandlikelybruisingfight.Butheandhis team had spent eighteen seemingly fruitless months working on Airbnb
beforeenteringYCombinator,rackinguptensofthousandsofdollarsincreditcarddebt.Afteralltheblood,sweat,andtears,weretheyreallywillingtogiveupaquarteroftheircompany?Ultimately,BriandecidednottobuyWimdu,swayedinpartbythearguments
ofhiskeyadvisers.FacebookfounderMarkZuckerbergcounseledhimtofight.“Don’tbuythem,”hesaid.“Thebestproductwillwin.”YC’s Paul Graham gave similar feedback. “They’re mercenaries. You’re
missionaries,” he told Brian. “They’re like people raising a baby they don’tactuallywant.”WhenBrianreachedout tomeformyadviceon thesituation, I tooadvised
himnottobuyWimdu.Thekeyissuewasn’tthepriceanddilution,butthewayamergercouldposeimpedimentstospeedandsuccess.“Buying[Wimdu]addsasubstantialamountofintegrationrisk,whichtrippedupGrouponafterbuyingCityDeal,” I told him. “Merging company cultures and companymanagementcouldcreatepotentiallyfatalrisks,especiallyifitslowsusdown.WithAirbnb,wehaveabusinessthatisalreadybenefitingfromnetworkeffects.Wecanwin.”Istandbythatadvicetoday.In the end, Airbnb’s founders realized that they wanted to take on the
Samwers—andtheywantedtowin.Buthow?The key was an aggressive, all-out program of growth that we call
blitzscaling. Blitzscaling drives “lightning” growth by prioritizing speed overefficiency,eveninanenvironmentofuncertainty.It’sasetofspecificstrategiesandtacticsthatallowedAirbnbtobeattheSamwerbrothersattheirowngame.Just a few months later, determined to acquire the resources needed to
outscale the Samwers,Brian raised $112million in additional venture capital.Airbnbthenembarkedonanaggressiveinternationalexpansionplan,includingtheacquisitionofAccoleo,asmallerandmoreaffordableGermanAirbnbclone,thatallowedAirbnbtocompetedirectlywithWimduinitshomemarket.Bythespringof2012,Airbnbhadopenednineinternationaloffices,settingupshopinLondon,Hamburg,Berlin,Paris,Milan,Barcelona,Copenhagen,Moscow,andSãoPaulo.Bookingshadgrownten timessince thatpreviousFebruary,and inJune2012Airbnbannounceditstenmillionthbooking.“The Samwers gave us a gift,” Brian admitted many years later in our
Blitzscalingclass.“Theyforcedustoscalefasterthanweeverwouldhave.”Bychoosingtogrowatabreakneckpace,Airbnbhadachievedadominantpositioninitsmarket.DespitetheinitialadvantagesthattheBerlin-basedWimduhadin
human resources, financial capital, and European market knowledge, thetechniques thatBrianandhiscofounders implementedallowedAirbnb tomeetandultimatelydefeatitschallenger.
2010:SHENZHEN,CHINA,TENCENTHEADQUARTERS
AboutayearbeforeAirbnbembarkedonitsblitzscalingjourney,inadifferentCEO’s office on the other side of the world, the message that would changeeverythingarrivedinthemiddleofthenight.Itwasthefallof2010,andPonyMa(Chinesename:MaHuateng)wastrying
to figure out what came next for Tencent, the company he had run sincefounding it in1998withfourclassmatesfromShenzhenUniversity.Thanks toits core product, the QQ instant messaging service, which had 650 millionmonthlyactiveusers,TencenthadbecomeoneofChina’smostvaluableInternetcompanies with revenues of nearly $2 billion, amarket capitalization of over$33billion, andmore than ten thousand employees.However,QQwasnowamature desktopproduct basedon late-1990s technology, and its user base hadstopped growing. Its American counterpart, AOL Instant Messenger, wasalreadyinaswiftdecline.Mawasconvinced thatTencenthad todevelopanewbreakthroughproduct
for the emerging smartphone platform—or else. “Internet companies that canreactwillsurvive,”hesaid,“andthosewhocan’twilldie.”ThemessagePonyMareadthatnightwasfromoneofTencent’semployees,
Allen Zhang (Chinese name: Zhang Xiaolong), a fellow entrepreneur whosecompany,Foxmail,Tencenthadacquiredfiveyearsearlier.Zhangnowranthecompany’s Guangzhou R&D division, which was a two-hour drive fromTencent’sShenzhenheadquarters.HehadbeenmonitoringtherapidgrowthofanewsocialmessagingproductcalledKik,whichwasespeciallypopularamongyoung people. He decided that Tencent needed to create its own socialmessengerforsmartphones—andquickly.Zhang’sproposalrepresentednotonlyahugeopportunitybutalsoahugerisk,
with equally huge uncertainty about the outcome. While a new messengerservicemightappealtoyoungconsumers,itwasprobablygoingtocannibalizeQQ, which was, after all, Tencent’s core business. Furthermore, Tencent hadpartneredwithleadingmobilecarrierslikeChinaMobiletoreceive40percentoftheSMScharges thatQQusers rackedupwhen they sentmessages tomobile
phones. A new service could hurt Tencent’s financial bottom line and at thesametimeriskitsrelationshipswithsomeofChina’smostpowerfulcompanies.It was the sort of decision that publicly traded, ten-thousand-person
companies typically refer to a committee for further study. But Ma wasn’t atypical corporate executive. That very night, he gave Zhang the go-ahead topursue the idea. Zhang put together a ten-person team, including sevenengineers,tobuildandlaunchthenewproduct.In just two months, Zhang’s small team had built a mobile-first social
messagingnetworkwithaclean,minimalisticdesignthatwasthepolaroppositeof QQ. Ma named the service Weixin, which means “micromessage” inMandarin.OutsideofChina,theservicebecameknownasWeChat.What came next was staggering. Just sixteen months after Zhang’s fateful
late-nightmessagetoMa,WeChatcelebrateditsonehundredmillionthuser.Sixmonthsafterthat,ithadgrowntotwohundredmillionusers.Fourmonthsafterthat,ithadgrowntothreehundredmillionusers.Pony Ma’s late-night bet paid off handsomely. Tencent reported 2016
revenues of $22 billion, up 48 percent from the previous year, and up nearly700 percent since 2010, the year before WeChat’s launch. By early 2018,Tencent reachedamarketcapitalizationofover$500billion,making itoneoftheworld’smostvaluablecompanies,andWeChatwasoneofthemostwidelyandintensivelyusedservicesintheworld.FastCompanycalledWeChat“China’sappforeverything,”andtheFinancial
Timesreportedthatmorethanhalfofitsusersspendoverninetyminutesadayusing the app. To put WeChat in an American context, it’s as if one singleservicecombined the functionsofFacebook,WhatsApp,FacebookMessenger,Venmo, Grubhub, Amazon, Uber, Apple Pay, Gmail, and even Slack into asingle megaservice. You can use WeChat to do run-of-the-mill things liketextingandcallingpeople,participatinginsocialmedia,andreadingarticles,butyoucanalsobookataxi,buymovietickets,makedoctors’appointments,sendmoneytofriends,playgames,payyourrent,orderdinnerforthenight,plussomuchmore.Allfromasingleapponyoursmartphone.Mahimselfrecognizedtheimportanceofthedecisionhehadmade,sayingin
aninterview,“Lookingback,thosetwomonthswereamatteroflifeanddeath.”Thesestoriesofextremegrowth,whetherinCaliforniaorhalfwayaroundthe
world in China, are perfect examples of why it’s valuable to study whatblitzscalingisandhowitworks.
Throughout this book, we will be telling the stories of various blitzscalers.Appendix B: The Blitzscalers includes brief profiles of these companies thatprovidemorecontext.Forevenmorebackground,visitBlitzscaling.com.
BLITZSCALING:THESECRETWEAPONFORBUILDINGSCALE-UPS
Whena start-upmatures to thepointwhere ithasakillerproduct, aclearandsizable market, and a robust distribution channel, it has the opportunity tobecomea“scale-up,”whichisaworld-changingcompanythattouchesmillionsorevenbillionsoflives.Often,thefastestandmostdirectpathfromstart-uptoscale-upisthehypergrowthproducedbyblitzscaling.TheenterprisesoftwarecompanySlackreachedthiscriticalstageonceitwas
able to demonstrate the rapid and accelerating adoption of its teammessagingapps by its initial market of software development teams. Nearly five yearspassedbetween the timewhenSlackwas founded and the initial launchof itsproduct. But once it launched, Slack users themselves drove user growth byaddingmanycolleagues at a time, aidedby a frictionlessprocess that allowednewuserstojumpinwithasimpleWebapplicationorbydownloadingamobileappfromiTunesorGooglePlay.Afterthecompanyreachedthispoint,itbeganto scale rapidly,addingemployees,capital, andcustomersatablisteringpace.Slackhad raised$17millionduring the first fiveyearsof its life;withineightmonthsoflaunch,ithadraisedanother$163millionandatotalof$800millionbylate2017.Anycompany,whether aglobal giant or a start-up in a cofounder’s garage,
would love to launch and grow killer businesses like Airbnb, WeChat, andSlack. Yet those who actually manage to do so, especially to the degree thatBrianChesky andPonyMadid, are still exceedingly rare.Why is that?Whatsetsthesecompaniesapartfromtherest?Inthisbook,wewillarguethatthekeytorapidlybuildingmassivebusinesses
intoday’senvironmentistheaggressivegrowthstrategyofblitzscaling:asetoftechniques that allows both start-ups and established companies to builddominant,world-leadingbusinessesinrecordtime.
ENTERINGTHEBLITZSCALINGERA
Overthepasttwodecades,theInternethascompletelyreshapedbothourdailylivesandtheworldofbusiness.Netscape’sblockbusterIPOonAugust9,1995,markedthebeginningofboththedot-comboomandwhatIcalltheNetworkedAge.Atthetime,therisingstockpricesofthedot-comboomattractedthemostattention, but, in retrospect, the biggest change was that the Internet wasbeginning to connect all of us to people, information, resources, and othernetworks.Therehavebeenotherrevolutionsinthepast—steam,electricity,andradiospringtomind—butwhatmakestheimpactoftheInternetsouniqueandsofar-reachingis thefact ithasmadeeverythingsomuchfaster.Today,everyindividual can connect to any other individual immediately; that increasedvelocityiswhatmakesblitzscalingpossibleandsopowerful.ThespeedoftheInternethasgeneratedanumberofsecond-ordereffectsthat
have changed how businesses and organizations can grow. For example, theInternet hasmade it possible to access globalmarkets and tap intomassivelyscalabledistributionchannels inaway thatwasn’t feasibleduringearlier eras.But perhaps the most important impact for businesses has been the risingsignificance and prevalence of so-called network effects that occur whenincreased usage of a product or service boosts the value of that product orservice for other users. For example, each additional Airbnb host makes theservice a tiny bit more valuable for every other Airbnb guest and vice versa.Each additional WeChat user makes the service a tiny bit more valuable foreveryotherWeChatuser,andsoon.Network effects generate a positive feedback loop that can allow the first
product or service that taps into those effects to build an unassailablecompetitive advantage. For example, eBaywas founded in 1995, yet networkeffects keep it a dominant player in peer-to-peer commerce twodecades later.Airbnboffersoverthreemillionlistingsinsixty-fivethousandcitiesaroundtheworld; think of howdifficult itwould be for a new entrant to offer anywhereclosetothesameselectionandvalue.We’reremindedofthefamousscenefromthemovieGlengarryGlenRoss,in
whichAlecBaldwin’scharacter,Blake,isspeakingtoagroupofsalesmen:
Asyouallknow,firstprizeisaCadillacEldorado.Anyonewannaseesecondprize?Secondprizeisasetofsteakknives.Thirdprizeisyou’refired.Getthepicture?
FirstprizeinthefirstwaveofconsumersocialnetworkingwenttoFacebook;secondprizetoMySpace;thirdprizetoFriendster.RememberFriendster?YouneedtowinfirstprizeinordertosurviveintheInternetera.Thelevelofcompetitioncanseemoverwhelmingattimes,buttheNetworked
Agealsoallowscompaniestoreapincrediblerewardsmuchmorerapidlythanatanyotherpointinhistory.Wecallthestrategyandmindsettheycanusetogetthere“blitzscaling.”Blitzscaling is a strategy and set of techniques for driving and managing
extremelyrapidgrowththatprioritizespeedoverefficiencyinanenvironmentofuncertainty. Put another way, it’s an accelerant that allows your company togrowatafuriouspacethatknocksthecompetitionoutofthewater.Blitzscalingrequireshypergrowthbutgoesbeyondthebluntstrategyof“get
big fast” because it involves purposefully and intentionally doing things thatdon’tmakesenseaccordingto traditionalbusiness thinking.In theBlitzscalingEra,youhavetomakeatoughcall:
Takeontheadditionalriskanddiscomfortofblitzscalingyourcompany,
Oracceptwhatmightbetheevengreaterriskoflosingifyourcompetitionblitzscalesbeforeyoudo.
WasAirbnb’sdecisiontoexpandintoEuropeanmarkets—amovethatcouldhave stretched the company so thin as to destroy its core business—eitherefficientorcertain?Hardly.Airbnbcouldeasilyhavefailed,burningthroughallitscapitalwhileessentiallycedingtheEuropeanmarkettoitscopycatcompetitorWimdu.Yettheriskydecisionprovedtherightone.Blitzscaling disrupts entire industries, such as music, video games, and
telephony,withbothnewtechnologiesandnewbusinessmodels…andthoseareexamples from just a single company. (You know, the one that produced theiPod, iTunes, the iPhone, and the iPad, to name just a few.) These waves ofdisruptionaffecteveryaspectofourdaily lives, fromthe jobswework, to theproductsweuse,tothewayweconnectwithoneanother.Disruptiononitsownisneithergoodnorbad,butitalwaysinvolveschange.
Replacinga$10productwitha$1productofequalorbetterqualitylookslikeadisaster to an incumbent player, but, for society as a whole, it means greaterproductivity.Thebuyergetsthedesiredproduct,andnowalsohas$9available
to invest in other things. Netflix has been bad news for broadcast and cablenetworks, but it has been great news for fans and creators of movies andtelevision.Yes,disruptionproduceslosersaswellaswinners,but,asawhole,itisavitalsourceofgrowthandopportunitythatyoucan’taffordtoignore.It’sgoodtokeepinmindthatthosewhoextollthevirtuesofdisruptiontend
to be—coincidentally enough—the ones in thewinners’ circle. But disruptionthat spreads its benefits and new opportunities broadly is better for society.Fortunately,most disruption falls into this category. In a 2004workingpaper,“SchumpeterianProfits in theAmericanEconomy:TheoryandMeasurement,”Yale economist William Nordhaus examined the US economy from 1948 to2001. Based on the data he collected, he concluded that only 2.2 percent of“profitsthatarisewhenfirmsareabletoappropriatethereturnsfrominnovativeactivity”went to the disrupters. “Most of the benefits of technological changearepassedon to consumers rather than capturedbyproducers,” he concluded.Likeitornot,changeisinevitable—butitdoesn’thavetobewhollyunexpected.In their bookFutureShock, the futuristsAlvin andHeidiTofflerwrote that
“change is the only constant,” and “to survive, to avertwhatwe have termedfutureshock,theindividualmustbecomeinfinitelymoreadaptableandcapablethaneverbefore.”Thosewordswereoriginallypublishedin1970.Thepaceofchangehasonlyacceleratedsincethen.Everyone should have the opportunity to learn how blitzscaling works,
because it isalready impacting their lives.Andonce theyknowhow itworks,theycanuseittoreshapetheworld.Peopleshouldbepartofbuildingthefutureratherthanfeelinglikethefutureisbeingforceduponthem.Blitzscalingiswhatseparatesthestart-upsthatgetdisruptedanddisappearas
theworld changes from the ones that scale up to becomemarket leaders andshapethefuture.ThisbookwasbornoutofaclasswetaughtatStanfordinwhichwedissected
theprocessthatwentintogrowingtheworld’slargesttechnologycompaniesandthencodifiedaseriesoftacticsandchoicesthatmadeitwork.Theresultwasaspecific set of principles that describes how to grow multibillion-dollarcompaniesinahandfulofyears.Whilewriting thisbook,we talked tohundredsof entrepreneurs andCEOs,
including those of the world’s most valuable companies, such as Facebook,Alphabet (Google), Netflix, Dropbox, Twitter, and Airbnb. (You can hear anumberof theseconversationsonmypodcast,MastersofScale.)Even though
the storiesof their companies’ risewereverydifferent inmanyways, theonethingtheyallhadincommonwasanextreme,unwieldy,risky, inefficient,do-or-dieapproachtogrowth.Inthisbook,wedrawlessonsfromtheseworld-leadingcompaniestoexplain
thenutsandboltsofhowtoblitzscale,whentoblitzscale,whytoblitzscale,andtheglobalimpactofthecompaniesthatareblitzscalingallaroundyourightthissecond.Thisquestwill takeusallover theglobe,butoneplace inparticular stands
out.
SILICONVALLEY:THEPERFECTPLACETODECODEBLITZSCALING
AlthoughcompanieshavesuccessfullyblitzscaledoneverycontinentexceptforAntarctica, themost prominent andmost concentrated set of examples comesfromCalifornia’sSiliconValley.Andwhilewecan’tsimplycopyandpastethetechniquesthatworkinSiliconValleyandexpectthemtoworkthesamewayinShanghai,neithercanwecutandpaste fromShanghai toStockholm,nor fromStockholm toSãoPaulo. Instead,we try toextract someuniversal lessonsandthenexplorehowtheyapplyacrosstheworld.Asofthiswritingattheendof2017,thereareonlyfourteenpubliclytraded
technology companies in the world that have a market capitalization of over$100billion.WanttoguesshowmanyofthoseareinSiliconValley?Seven—that’shalfoftheworld’smostvaluabletechcompanies.Taken together, Silicon Valley’s 150 most valuable publicly traded
technologycompaniesareworth$3.5 trillion.Thatnumber issobig itdoesn’tmeananythingtomostofus.Soconsiderthis:those150companiesalonemakeup50percentofthevalueoftheNASDAQ,andtheyaccountforover5percentof the entire world’smarket capitalization. That’s a lot of value created by aregionwithanestimated3.5 to4million residents,or roughly0.05percentoftheworld’spopulation.Whilewe fully accept that thismay change in the future, the historical and
current success of Silicon Valley makes it the perfect place to examine thisquestion:What is the most effective way to rapidly build massively valuablecompanies?When outsiders look at SiliconValley, they often think that the key to this
question is innovative technology.Butasyou’ll read, technological innovationalonedoesn’tmakeforathrivingcompany.Silicon Valley insiders and well-read outsiders believe that the key is the
combinationof talent,capital,andentrepreneurialculture thatmakes iteasy tostartnewcompanies.Thistooiswrong.Sure,SiliconValleyistheleadinghubforhigh-techtalentandventurecapital,
butitdidn’tstartoutthatway.Sure,itisblessedwithgreatuniversities,suchasStanfordandBerkeley,butsoareplentyofotherregions.Theanswercan’tbesimply the combination of venture capital, research universities, and smartpeople. This combination of ingredients is far from unique. In fact, the samebasicingredientscaneasilybefoundinnumerousstart-upclustersintheUnitedStates and around the world: Austin, Boston, New York, Seattle, Shanghai,Bangalore,Istanbul,Stockholm,TelAviv,andDubai.Todiscover the secret toSiliconValley’s success,youneed to lookbeyond
thestandardoriginstory.Whenpeople thinkofSiliconValley, the first thingsthatspring tomind—after theHBOtelevisionshow,ofcourse—are thenamesof famous start-ups and their equally glamorized founders: Apple, Google,Facebook;Jobs/Wozniak,Page/Brin,Zuckerberg.The success narrative of these hallowed names has become so universally
familiar that people fromcountries around theworld can tell it just aswell asSand Hill Road venture capitalists. It goes something like this: A brilliantentrepreneurdiscoversanincredibleopportunity.Afterdroppingoutofcollege,heorshegathersasmallteamwhoarehappytoworkforequity,setsupshopinahumblegarage,playsfoosball,raisesmoneyfromsageventurecapitalists,andproceeds to change theworld—afterwhich, of course, the founders and earlyemployeeslivehappilyeverafter,usingthewealththey’veamassedtofundbothanewgenerationofentrepreneursandasetofeponymousbuildingsforStanfordUniversity’sComputerScienceDepartment.It’s an exciting and inspiring story. We get the appeal. There’s only one
problem.It’sincompleteanddeceptiveinseveralimportantways.First,while “SiliconValley” and “start-ups” are used almost synonymously
these days, only a tiny fraction of the world’s start-ups actually originate inSiliconValley,andthisfractionhasbeengettingsmallerasstart-upknowledgespreadsaroundtheglobe.ThankstotheInternet,entrepreneurseverywherehaveaccesstothesameinformation.Moreover,asothermarketshavematured,smartfoundersfromaroundtheglobeareelectingtobuildcompaniesinstart-uphubs
intheirhomecountriesratherthanimmigratingtoSiliconValley.Second,simplystartingacompanyisobviouslyinsufficient.Thestart-upsthat
achievemassivevaluearethosethathavefoundawaytogrowintoscale-upsatanexponentiallyfasterpacethantheircompetitors.Sowhat secret alchemy is atwork in SiliconValley to fuel such rapid-fire
growthofsomanyoftheworld’smostvaluabletechcompanies?Andifthereisasecret,canitbeidentified,analyzed,understood,and,mostimportant,appliedelsewhere?Blitzscalingisthatsecret.Andthereasonblitzscalingmatterssomuchisthat
nothingaboutitisinherenttoSiliconValley.There’sacommonmisconceptionthatSiliconValleyistheacceleratorofthe
world.Therealstoryisthattheworldkeepsgettingfaster—SiliconValleyisjustthefirstplacetofigureouthowtokeeppace.WhileSiliconValleycertainlyhasmany key networks and resources thatmake it easier to apply the techniqueswe’regoingtolayoutforyou,blitzscalingismadeupofbasicprinciplesthatdonotdependongeography.We’regoingtoshowyouexamplesfromoverlookedpartsof theUnitedStates, suchasDetroit (RocketMortgage)andConnecticut(Priceline), as well as from international companies, such as WeChat andSpotify.Intheprocessyou’llseehowthelessonsofblitzscalingcanbeadaptedto help build great companies in nearly any ecosystem, albeit with differingdegreesofdifficulty.That’sthemissionofthisbook.Wewanttosharethesecretweaponthathas
allowedSiliconValleytopunchsomuch(morethanahundredtimes)aboveitspopulationindexsothatthoselessonscanbeappliedfarbeyondthesixty-milestretchbetweentheGoldenGateBridgeandSanJose.Itissorelyneeded.Here’s a startling fact: the global economywill need to create six hundred
millionnewjobsby2030tomeettheUnitedNations’sustainabledevelopmentgoals.That’slessthanfifteenyearsaway.Theworldneedsmorethanjustnewcompaniesandnewjobs;it’sgoingtoneedentirenewindustries.Thoseindustriesbettergeneratescale-upsaswellasstart-ups.Itseemstous
that itwill be a lot easier to add six hundredmillion new jobsworldwide bycreating sixty thousand new ten-thousand-person companies rather than sixtymillionnewten-personcompanies.Thelate,greatAndyGrove,Intel’slegendaryCEO,understoodandexplained
thiswhenhewroteina2010op-edforBloomberg:
Start-ups are a wonderful thing, but they cannot by themselvesincrease tech employment. Equally important iswhat comes afterthatmythicalmomentofcreationinthegarage,astechnologygoesfrom prototype to mass production. This is the phase wherecompaniesscaleup.Theyworkoutdesigndetails,figureouthowtomake things affordably, build factories, and hire people by thethousands.Scaling ishardworkbutnecessary tomake innovationmatter.
Recognizing what powers the rapid growth from start-up to scale-up, andunderstanding the principles behind how itworks,will help entrepreneurs andcompaniesapply theseprinciplesnot just insmallpocketsof theUnitedStatesandChinabutaroundtheworld.
WHOSHOULDREADTHISBOOK?
This book is for anyonewhowants to understand the techniques that allow abusinesstogrowfromzerotoamultibillion-dollarmarketleaderinahandfulofyears.These techniques should be of interest to entrepreneurs who want to build
massivecompanies,venturecapitalistswhowant to invest in them,employeeswhowant to work for them, and governments and communities who wish toencouragethegrowthofthesecompaniesintheirownregions.Andevenifyoudon’twant tobuild, invest in, orwork for anyof these companies,you’ll stillneedtonavigatetheworldthatthey’rebuilding.Ifyouareamanageroraleaderwhoistryingtorapidlyscaleaprojectora
businessunitwithinalargercompany,blitzscalingcanhelpyoutoo.Andwhilewe draw these lessons primarily from the world of high tech, many of theprinciples and frameworks the book lays out (especially regarding peoplemanagement) are applicable to high-growth companies in most industriesworldwide,fromEuropeanfast-fashionretailerstoTexanoilshalecompanies.Even organizations outside the business world can use blitzscaling to their
advantage. Upstart presidential campaigns and nonprofits serving theunderprivileged have used the levers of blitzscaling to overturn conventionalwisdom and achieve massive results. You’ll read all these stories, and many
more,inthepagesofthisbook.Whetheryouareafounder,amanager,apotentialemployee,oraninvestor,
we believe that understanding blitzscaling will allow you to make betterdecisionsinaworldwherespeedisthecriticalcompetitiveadvantage.Withthepowerofblitzscaling,theadoptedsonofaSyrianimmigrant(Steve
Jobs),theadoptedsonofaCubanimmigrant(JeffBezos),andaformerEnglishteacherandvolunteertourguide(JackMa)wereallabletobuildbusinessesthatchanged—andarestillchanging—theworld.The strategy and techniques we describe in this book are based on my
experiences as a member of the founding team at PayPal; as the cofounder,CEO, and now executive chairman at LinkedIn; as a leading investor inFacebookandAirbnb;andasaninvestoratGreylockPartners,whereIworkedwithmanyotherbillion-dollarcompanies,suchasWorkday,Pandora,Cloudera,andPureStorage.MypartnersatGreylockandIhavehelped thesecompaniesgo from garage to global dominance, and, in this book, we’ll share with youwhatwebelieveareimportantframeworksforunderstandingandaddressingthechallengeofblitzscalingacrossthedifferentelementsofyourorganization.Yet asmany good business books disclaim,while this is a playbook and a
strategyguide,itisn’tabookofpreciserecipes.Regardlessofhowthepopularpressportraysthings,eachformulaforbuildingagreatcompanyisuniqueanddependsonthemarketopportunity,thefounders,andthenetworkinwhichtheyoperate.Thetruthisthereisabsolutelynothingguaranteedasaone-size-fits-all,must-followrulebookforeveryone.However,therearepatterns.Soinadditiontoindividualtipsandtricks,thisbookoffersasetofframeworksandstrategiesfor leaders, entrepreneurs, and intrapreneurs to adapt to their own needs andcircumstances.
AQUICKNOTEONTHETERM“BLITZSCALING”
Theterm“blitzscaling”derivesfromthetwentieth-centuryusageof“blitz”asawayof describing a sudden, all-out effort.The first usage of blitz in thiswaywas to describe the “blitzkrieg” (“lightningwar”) strategy thatGeneralHeinzGuderian devised for the initial military campaigns of Nazi Germany duringWorldWar II. Ironically enough, Guderianwas heavily influenced by Britishmilitary thinkers like Basil Liddell Hart and J. F. C. Fuller, and the term“blitzkrieg”wasactuallypopularizedby theBritishpress; theGermanmilitary
neverformallyadoptedit.Theadvancingarmiesinthesecampaignsabandonedthetraditionalapproach
ofmovingattheslowpaceatwhichtheycouldestablishsecurelinesofsupplyandretreat. Instead theyfullycommitted toanoffensivestrategy thatacceptedthe possibility of running out of fuel, provisions, and ammunition, riskingpotentiallydisastrousdefeatinordertomaximizespeedandsurprise.Thespeedof these armies’ advance shocked and overwhelmed their opponents, allowingtheblitzkriegerstooutmaneuverandoutfightthedefendingforces.TheearlysuccessoftheGermanarmyhelpedspreadthelessonsofblitzkrieg
toalltheforcesinthewar.Forexample,theAmericangeneralGeorgeS.PattonlaterputtheselessonstogooduseinleadingtheUSThirdArmy’sadvancefromthebeachesofNormandyallthewaytoBerlin.Sincethen,theterm“blitz”hasbeenusedtodescribeeverythingfromanAmericanfootballplaytothewayinwhichlargecorporationsrolloutnewproducts.Liketheall-outblitzdefenseinfootball—whichinvolvestheriskymoveofsendingeveryavailabledefendertopursue the quarterback—or the proverbial marketing blitz of television, print,andonlineadvertisingthataccompaniesthereleaseofanewblockbustermovie,blitzscaling strives for a relentless and dizzying speed that overwhelms themarket.Whilewearewaryofthenegativeconnotationsof“blitz,”especiallyinthose
nations that felt the effects of blitzkrieg inWorldWar II,we believe that thestrengthof themetaphor and thewidespreadandcolloquialuseof the term innonmilitarycontextsmakeitthebestfitfortheconceptsdiscussedinthisbook.
PARTI
WhatIsBlitzscaling?
Blitzscaling is what we call both the general framework and the specifictechniquesthatallowcompaniestoachievemassivescaleatincrediblespeed.Ifyou’re growing at a rate that is so much faster than your competitors that itmakesyoufeeluncomfortable,thenholdontight,youmightbeblitzscaling!Amazon’s incredible growth in the late 1990s (and up through today) is a
prime example of blitzscaling. In 1996, a pre-IPO Amazon Books had 151employees and generated revenues of $5.1 million. By 1999, the now-publicAmazon.com had grown to 7,600 employees and generated revenues of$1.64 billion. That’s a 50 times increase in staff and a 322 times increase inrevenue in just threeyears. In2017,Amazonhad541,900employeesandwasforecasttogeneraterevenuesof$177billion(upfrom$136billionin2016).DropboxfounderDrewHoustondescribed thefeelingproducedby thiskind
of growthwhenhe toldme, “It’s like harpooning awhale.Thegoodnews is,you’veharpoonedawhale.Andthebadnewsis,you’veharpoonedawhale!”While blitzscaling may seem desirable, it is also fraught with challenges.
Blitzscalingisjustaboutascounterintuitiveasitcomes.Theclassicapproachtobusiness strategy involves gathering information and making decisions whenyoucanbereasonablyconfidentoftheresults.Takerisks,conventionalwisdomsays,buttakecalculatedonesthatyoucanbothmeasureandafford.Implicitly,thistechniqueprioritizescorrectnessandefficiencyoverspeed.Unfortunately, this cautious and measured approach falls apart when new
technologiesenableanewmarketorscrambleanexistingone.Chris earned his MBA from Harvard Business School in the late 1990s,
duringthedawnoftheNetworkedAge.Backthen,hisMBAtrainingfocusedontraditional techniques, such as using discounted cash flow analysis to makefinancial decisions with greater certainty. Chris also learned about traditionalmanufacturing techniques, such as how to maximize the throughput of anassemblyline.Thesemethodsfocusedonachievingefficiencyandcertainty,andthe same emphasis was reflected in the broader business world. The world’smostvaluablecompanyduringthattime,GeneralElectric,wasbelovedbyWallStreet analysts for its ability to deliver consistent and predictable earningsgrowth. But efficiency and certainty, while innately appealing, and veryimportant in thecontextofastable,establishedmarket,offer littleguidance tothedisrupters,inventors,andinnovatorsoftheworld.Whenamarketisupforgrabs,theriskisn’tinefficiency—theriskisplaying
it toosafe. Ifyouwin,efficiencyisn’t that important; ifyoulose,efficiencyiscompletelyirrelevant.Overtheyears,manyhavecriticizedAmazonforitsriskystrategyofconsumingcapitalwithoutdeliveringconsistentprofits,butAmazonisprobablygladthatits“inefficiency”helpeditwinseveralkeymarkets—onlineretail,ebooks,andcloudcomputing,tonamejustafew.When you blitzscale, you deliberately make decisions and commit to them
eventhoughyourconfidencelevelissubstantiallylowerthan100percent.Youaccept the risk of making the wrong decision and willingly pay the cost ofsignificant operating inefficiencies in exchange for the ability to move faster.Theserisksandcostsareacceptablebecausetheriskandcostofbeingtooslowisevengreater.Butblitzscaling ismore than justplungingaheadblindly inaneffortto“getbigfast”towinthemarket.Tomitigatethedownsideoftherisksyou take,you should try to focus them—line themupwitha smallnumberofhypotheses about howyour businesswill develop so that you canmore easilyunderstandandmonitorwhatdrivesyoursuccessorfailure.Youalsohavetobepreparedtoexecutewithmorethan100percentefforttocompensateforthebetsthatdon’tgoyourway.Forexample,anyonewhoknowsJeffBezosknowsthathedidn’tsimplymash
hisfootdownonthegaspedal;Amazonhasintentionallyinvestedaggressivelyin the future, and, despite its accounting losses, generates a ton of cash.Amazon’s operating cash flow was over $16 billion in 2016, but it spent$10 billion in investments and $4 billion paying down debt. Its seeminglymeagerprofitsareafeatureofitsaggressivestrategy,notabug.Blitzscaling requires more than just courage and skill on the part of the
entrepreneur.Italsorequiresanenvironmentthatiswillingtofinanceintelligentrisks with both financial capital and human capital, which are the essentialingredientsforblitzscaling.Thinkofthemasfuelandoxygen;youneedbothtopropeltherocketskyward.Meanwhile,theinfrastructureofyourorganizationisthe actual structure of your rocket,which you’re rebuilding on the fly as yourise. Your job as a leader and an entrepreneur is tomake sure that you havesufficient fuel to propel your growth while making the necessary mechanicaladjustmentstotheactualrocketshiptokeepitfromflyingapartasitaccelerates.Fortunately,thisismorepossibletodaythanithaseverbeeninthepast.
SOFTWAREISEATING(ANDSAVING)THEWORLD
Historically, stories of breakneck growth involved either computer software,which offers nearly unlimited scalability in terms of distribution, or software-enabledhardware,suchastheFitbitfitnesstrackerorTeslaelectriccar,whosesoftware component allows the company to innovate on software timescales(days orweeks) rather than hardware timescales (years).Moreover, the speedandflexibilityofsoftwaredevelopmentallowcompanies to iterateandrecoverfromtheinevitablemisstepsofhaste.What’s especially exciting these days is that software and software-enabled
companies are starting to dominate industries outside of traditional high tech.My friend Marc Andreessen has argued that “software is eating the world.”Whathemeans is that even industries that focusonphysicalproducts (atoms)are integrating with software (bits). Tesla makes cars (atoms), but a softwareupdate (bits) can upgrade the acceleration of those cars and add an autopilotovernight.The spread of software and computing into every industry, along with the
dense networks that connect us all, means that the lessons of blitzscaling arebecomingmore relevant and easier to implement, even inmature or low-techindustries.Touseacomputingmetaphor,technologyisacceleratingtheworld’s“clock speed” (the rate at which Central Processing Units [CPUs] operate),making change occur faster than previously thought possible. Not only is theworldmovingfaster,butthespeedatwhichmajornewtechnologyplatformsarebeing created is reducing the downtime between the arrivals of eachwave ofinnovation.Before,individualwaveswouldsweepthroughtheeconomyoneatatime—technologieslikepersonalcomputers,diskdrives,andCD-ROMs.Today,
multiplemajorwavesseemtobearrivingsimultaneously—technologieslikethecloud,AI,AR/VR,nottomentionmoreesotericprojectslikesupersonicplanesand hyperloops. What’s more, rather than being concentrated narrowly in apersonal computer industry that was essentially a niche market, today’s newtechnologies impact nearly every part of the economy, creating many newopportunities.Thistrendholdstremendouspromise.Precisionmedicinewillusecomputing
power to revolutionize health care. Smart grids use software to dramaticallyimprove power efficiency and enable the spread of renewable energy sourceslikesolarroofs.Andcomputationalbiologymightallowustoimprovelifeitself.Blitzscaling can help these advances spread and magnify their sorely neededimpact.
THETYPESOFSCALING
Blitzscaling isn’t simplyamatterof rapidgrowth.Every company isobsessedwithgrowth.Inanyindustry,youliveanddiebythenumbers—useracquisition,margins, growth rate, and so on.Yet growth alone is not blitzscaling.Rather,blitzscaling isprioritizing speedover efficiency in the face of uncertainty.Wecan better understand blitzscaling by comparing it to other forms of rapidgrowth.
Efficiency Speed
Uncertainty ClassicStart-upGrowth Blitzscaling
Certainty ClassicScale-upGrowth Fastscaling
Classic start-up growth prioritizes efficiency in the face of uncertainty.Startingacompanyislikejumpingoffacliffandassemblinganairplaneontheway down; being resource-efficient lets you “glide” to minimize the rate ofdescent,givingyouthetimetolearnthingsaboutyourmarket,technology,andteambeforeyouhittheground.Thiskindofcontrolled,efficientgrowthreducesuncertainty and is a good strategy to follow while you’re trying to establishcertaintyaroundwhattheauthorsEricRiesandSteveBlankcallproduct/marketfit:yourproductsatisfiesastrongmarketdemandfor thesolutiontoaspecificproblemorneed.Classic scale-up growth focuses on growing efficiently once the company
has achieved certainty about the environment. This approach reflects classiccorporatemanagement techniques, such as applying “hurdle rates” so that thereturnoninvestment(ROI)ofcorporateprojectsconsistentlyexceedsthecostofcapital.Thiskindofoptimizationisagoodstrategytofollowwhenyou’retryingtomaximizereturnsinanestablished,stablemarket.Fastscalingmeans that you’rewilling to sacrifice efficiency for the sakeof
increasing your growth rate. However, because fastscaling takes place in anenvironment of certainty, the costs are well understood and predictable.Fastscaling is a good strategy for gaining market share or trying to achieverevenue milestones. Indeed, the financial services industry is often happy tofinance fastscaling, whether by buying stocks and bonds or lending money.Analysts and bankers feel confident that they can create elaborate financialmodelsthatworkouttothepennythelikelyROIofafastscalinginvestment.Blitzscalingmeans that you’rewilling to sacrifice efficiency for speed, but
without waiting to achieve certainty on whether the sacrifice will pay off. Ifclassic start-up growth is about slowing your rate of descent as you try toassemble your plane, blitzscaling is about assembling that plane faster, thenstrapping on and igniting a set of jet engines (and possibly their afterburners)while you’re still building the wings. It’s “do or die,” with either success ordeathoccurringinaremarkablyshorttime.Given these definitions, youmight wonder why anyone would ever pursue
blitzscaling. After all, it combines the gut-wrenching uncertainty of start-upgrowth with the potential for a much bigger, more embarrassing, moreconsequentialfailure.Blitzscalingisalsohardtoimplement.Unlessyou’relikeMicrosoft or Google and can finance your growth from an exponentiallygrowingrevenuestream,you’llneed toconvince investors togiveyoumoney,and it’s much harder to raise money from investors for a calculated gamble(blitzscaling) than for a sure thing (fastscaling). To make matters worse, youusually needmore money to blitzscale than to fastscale, because you have tokeepenoughcapitalinreservetorecoverfromthemanymistakesyou’relikelytomakealongtheway.Yetdespiteallofthesepotentialpitfalls,blitzscalingremainsapowerfultool
forentrepreneursandotherbusinessleaders.Ifyou’rewillingtoaccepttherisksofblitzscalingwhenothersaren’t,you’llbeabletomovefasterthantheywill.Ifthe prize to be won is big enough, and the competition to win it is intenseenough,blitzscalingbecomesarational,evenoptimalstrategy.
Once you convince themarket for capital and themarket for talent—whichincludeclients andpartners, aswell as employees—to invest inyour scale-up,you have the fuel required to start blitzscaling. At that point, your objectiveswitches from going from zero to one to going from one to one billion in anincrediblycompressedtimeframe.Acompanymightemploydifferent typesofscalingatdifferentpoints in its
life cycle. The canonical sequence that companies like Google and Facebookhave gone through begins with classic start-up growth while establishingproduct/market fit, then shifts into blitzscaling to achieve criticalmass and/ormarketdominanceaheadofthecompetition,thenrelaxesdowntofastscalingasthebusinessmatures,andfinallydownshiftstoclassicscale-upgrowthwhenthecompany is an established industry leader. Together, this sequence of scalinggeneratesaclassic“S-curve”ofgrowth,withslowerinitialgrowthfollowedbyrapidacceleration,eventuallyeasingitswayintoagentleplateau.
Of course, this canonical sequence is greatly simplified. The scaling cycleapplies not just to whole companies but to individual products and businesslines; the aggregate curves of these scaling cycles generate the overall scalingcurveforthecompany.For example,Facebookbeganas a classicblitzscaling story.Theyear-over-
yearrevenuegrowthduringitsfirstfewyearsofexistencewere2,150percent,433 percent, and 219 percent, going from zero to $153million in revenue in2007.Thenthecompanywentthroughakeytransition,andgrowthdroppedintothe double-digit range as Facebook struggled with both monetization and theshift fromdesktop tomobile.Fortunately,Facebook founderMarkZuckerbergmade two important moves: he personally led a shift from desktop-first tomobile-first,andhehiredSherylSandbergasthecompany’sCOO,whointurn
builtFacebook intoanadvertising sales juggernaut.Growth roseback into thetriple-digitrange,and,by2010,thesemoveshadpushedFacebook’srevenuestoover$2billion.We’llexaminebothofthesekeymovesingreaterdetaillaterinthebook,withFacebook’sshifttomobilefeaturedinouranalysisofFacebook’sbusinessmodel,andFacebook’shiringofSherylSandberginthesectiononthekeytransitionfromcontributorstomanagerstoexecutives.Apple illustrateshowthisoverlap looksovermultipledecades. In itsstoried
history, Apple went through complete scaling cycles for the Apple II, theMacintosh, the iMac, and the iPod (with the cycle for the iPhone still underway). It’s worth noting that Apple failed to launch any blitzscalable productsaftertheAppleIIandtheMacuntilSteveJobsreturnedandlaunchedtheiMac,iPod,andiPhone.ItwaspartofSteve’sraregeniusthattimeandtimeagainhewasabletopicktherightproductforAppletoblitzscale,evenwithoutslowingdownforaperiodofclassicstart-upgrowthtogatherfeedbackfromthemarket.
The scaling curve applies to every blitzscaler, regardless of industry orgeography.ThesamemultipleS-curvegraphthatdescribesFacebookorApplealsodescribesTencent,whichlaunchedwithQQ,thenaddedasecondcurveforWeChat after QQ reached maturity in 2010. Just when you’ve finishedblitzscalingonebusiness line,youneed toblitzscale thenext tomaintainyourcompany’s upward trajectory. And as blitzscaling continues to spread,established companies with mature business lines should consider turning tointrapreneurstoblitzscalenewbusinessunits.
THETHREEBASICSOFBLITZSCALING
Blitzscaling requires you to move at a pace that is almost certainly
uncomfortable for your team.Youwill definitelymakemanymistakes as younavigateanenvironmentfullofuncertainty;theartliesindevelopingtheskilltolearnquicklyfromthosemistakesandreturntoarelentlesslyrapidadvance.Butfirst,it’scriticaltounderstandthreebasics.
1.BLITZSCALINGISBOTHANOFFENSIVESTRATEGYANDADEFENSIVESTRATEGY.
Onoffense,blitzscalingallowsyoutodoseveralthings.First,youcantakethemarket by surprise, bypassing heavily defended niches to exploit breakoutopportunities. For example, Slack’s rapid growth after its launch blindsided ahostofentrenchedcompetitorslikeMicrosoftandSalesforce.com.Second,youcan leverageyour lead tobuild long-termcompetitive advantagesbeforeotherplayersareabletorespond.We’llexplorethisconceptingreaterdetaillateron.Third,blitzscalingopensupaccesstocapital,becauseinvestorsgenerallypreferto backmarket leaders.You canwin thismantle if youblitzscale, andwith itraisemoremoneymoreeasilyandmorequicklythanyourlaggingcompetitors.On defense, blitzscaling lets you set a pace that keeps your competitors
gaspingsimplytokeepup,affordingthemlittletimeandspacetocounterattack.Because they’re focused on responding to your moves, which can often takethembysurpriseandforcethemtoplaycatch-up,theydon’thaveasmuchtimeavailable to develop and execute differentiated strategies that might threatenyour position.Blitzscaling helps youdetermine the playing field to your greatadvantage.
2.BLITZSCALINGTHRIVESONPOSITIVEFEEDBACKLOOPS,INTHATTHECOMPANYTHATGROWSTOSCALEFIRSTREAPSSIGNIFICANTCOMPETITIVEADVANTAGES.
InApril2014,McKinsey&Companypublishedareportentitled“Growfastordieslow,”whichanalyzedthelifecyclesofthreethousandsoftwareandInternetcompanies, and found thatpositive feedback loopsmade rapidgrowth thekeyfactorinfinancialsuccess:
First,growthyieldsgreaterreturns.High-growthcompaniesoffera
return to shareholders five times greater than medium-growthcompanies. Second, growth predicts long-term success.“Supergrowers”—companies whose growth was greater than60 percent when they reached $100 million in revenues—wereeight timesmore likely to reach$1billion in revenues than thosegrowinglessthan20percent.
We believe that the mechanism behind the power of blitzscaling is “first-scaler advantage.”Once a scale-upoccupies thehighground in its ecosystem,thenetworksarounditrecognizeitsleadership,andbothtalentandcapitalfloodin.For one, top professionals understand that they can have a greater impact
working for themarket leader.Meanwhile, joining a scale-up that is clearly a“rocketship”offersmanyofthefinancialrewardsofworkingforanearly-stagestart-up,with farmorecertaintyandfar less risk.Scale-upemployeesarepaidmarketsalaries,receiveequityupside,andhaveaverygoodchanceofbecomingrich, if not filthy rich. By attracting the best people, scale-ups increase theirabilitytobuildandbringtomarketgreatproducts,whichinturnincreasestheirabilitytorapidlyscale.A parallel calculus applies to investors. Venture capitalists (VCs) make
investment decisions based on the confidence interval they have in theirinvestmentthesis.Achievingscaleshrinksthoseintervalsandmakesiteasiertodecide to invest.And because the network that connects investors—especiallywithin a tight-knit ecosystem like Silicon Valley—can disseminate thisinformationquicklyandbroadly,ablitzscalingcompanycan raisecapitalonamassivescale.Thiscapitalinfusioncanfuelexplosivegrowth,whichshrinkstheconfidenceintervalsevenfurther.Paradoxically, globalization has both leveled the playing field for
entrepreneurs around theworldand increased the value of being in a premierscalinghublikeSiliconValleyorChina.Becausetherestoftheworldbelievesthattheseecosystemshaveanadvantageinscalingupstart-ups,thosestart-upsandtheirinvestorsattractcapital(humanandfinancial)fromallovertheworld,furtherbolsteringtheirabilitytokeepgrowing.Thisisakeyreasonwhyscale-upslikeUberandPinteresthaveachievedascaleandvaluationthatdwarfthoseofmostpubliclytradedcompanies.DuetomyroleatGreylockPartners,Ican’tcommenton thevaluationsofDropboxandAirbnb,but theyoccupya similar
placeintheecosystem.Consider thecaseof twovery similarcompanies,TwitterandTumblr.Both
hadbrilliant,product-orientedfoundersinEvan“Ev”WilliamsandDavidKarp.Both were hot social media start-ups. Both grew at a remarkable rate afterestablishingproduct/marketfit.Bothhadamajorimpactonpopularculture.YetTwitterwentpublicandachievedamarketcapitalization thatpeakedatnearly$37billion,whileTumblrwasacquiredbyYahoo!—anotherstart-up thatusedblitzscaling to become a scale-up, only to decline and fade away—for “only”$1billion.Was thisdumb luckonTwitter’s side?Perhaps.Luckalwaysplays a larger
role than founders, investors, and themediawould like to admit.But amajordifferencewas that Twitter could draw on numerous networks for advice andhelp that Tumblr could not. For example, Twitter was able to bring in DickCostolo,asavvyexecutivewithpriorscalingexperienceatGoogle.Incontrast,eventhoughTumblrwasarguablythemostprominentstart-upinitsNewYorkCity ecosystem, it couldn’t easily draw upon a pool of local talent who hadexperiencedealingwith rapidgrowth.According toGreylock’s JohnLilly, foreveryexecutiverolethatTumblrneededtofill,therewerelessthanahandfulofcandidatesinallofNewYorkCity.Thispaucityoftalentmadehiringdifficult;thecompanywasreluctanttoreplaceexistingemployeesduetoalackofbetteralternatives.Without theability tohireanexecutive teamthatcouldblitzscale,Tumblrdecidedtosellthecompany.Of course, while geography can present challenges to blitzscaling, they
becomemuchmoresolvableifyou’reawareofthem.Forexample,overthepastdecade, Priceline—the world’s most successful online travel company—hasbeenabletoblitzscalefromitsheadquartersinConnecticut.TheCEOwholedPriceline during its growth phase, Jeffery Boyd, saw advantages to thisgeographic isolation, noting that the company’s location meant that it facedfewer bidding wars for the key software engineers and designers needed tosupporttherapidgrowthofthebusiness.It’sextremelydifficultforlaterentrantstocompetedirectlywithablitzscaling
company that has first-scaler advantage. Unless these players find a differentgameinwhichtheycancapturethisadvantage,they’llsimplybecomeirrelevant.
3.DESPITEITSINCREDIBLEADVANTAGESANDPOTENTIAL
PAYOFFS,BLITZSCALINGALSOCOMESWITHMASSIVERISKS.
Untilrecently,“Movefastandbreakthings”wasFacebook’sfamousmotto.Yetrapid growth can cause nearly as many problems as it solves. As MarkZuckerbergtoldmeinaninterviewformyMastersofScalepodcast,“Wegottoapointwhereitwastakingusmoretimetogobackandfixthebugsandissuesthatwe’recreatingthanthespeedthatweweregainingbygoingfaster.”Inonefamousincident,asummerinternintroducedabugthatbroughtdowntheentireFacebooksiteforthirtyminutes.There is a scientific term for out-of-control growth in the human body:
“cancer.”Inthiscontext,uncontrolledgrowthisclearlyundesirable.Thesameistrueforabusiness.Successfulblitzscalingmeansthatyou’remaintainingatleastsomelevelofcontrolbyrapidlyfixingthethingsthatwillinevitablygetbrokenso that the company can maintain its furious pace without flaming out orcollapsinginonitself.LikeanAmericanfootballplayerstreakingdownthefieldfor a game-winning touchdown, even a company that has achieved first-scaleradvantagecanlosetheballpriortocrossingthegoallineifittakesonabiggerriskthanitcanhandle.Blitzscaling is risky from a management perspective as well. Reinventing
yourleadershipstyle,yourproduct,andyourorganizationateverynewphaseofscalewon’tbeeasy,butitisnecessary.InthewordsofleadershipguruMarshallGoldsmith,“Whatgotyouherewon’tgetyouthere.”Market share and revenue growth earn headlines, but you can’t achieve
customerandrevenuescalewithoutscalingupyourorganization,intermsofthesizeandscopeofyourstaff,aswellasyourfinancial,product,andtechnologystrategy. If the organization doesn’t grow in lockstep with its revenues andcustomerbase,thingscanquicklyspiraloutofcontrol.Forexample,duringaperiodofblitzscalinginthelate1980sandearly1990s,
Oracle Corporation focused so single-mindedly on sales growth that itsorganization lagged badly on both technology (where it fell behind archrivalSybase’s) and finance and nearly went bankrupt as a result. It took theturnaround efforts of Ray Lane and Jeff Henley to stave off disaster andrepositionOracleforitslatersuccess.Blitzscaling your organization will require hard choices and sacrifices; for
example, the peoplewho are adept at launching a company aren’t necessarilygoing to be the right people to scale it, as the Oracle example abovedemonstrates. Later in the book we’ll discuss how successful blitzscalers
consciouslymanagegrowthratherthanlettingitmanagethem.
THEFIVESTAGESOFBLITZSCALING
Blitzscalingastart-upisn’talinearprocess;aglobalgiantisn’tsimplyastart-upthat’s been multiplied by one thousand, working out of a gleaming high-riseheadquarters instead of a grimy garage. Each major increment of growthrepresents a qualitative as well as quantitative change. Drew Houston ofDropbox expressed thiswellwhen he toldme, “The chessboard keeps addingnewpiecesandnewdimensionsovertime.”In the physical sciences, materials often undergo phase changes as their
circumstances (e.g., temperature and pressure) change. Ice melts into water;water boils into steam.As a start-up scales up from one phase to the next, itundergoesfundamentalchangesaswell.Andinthesamewaythaticeskatesareuselessonwater,andyoucan’tskip
rocksonwatervapor, theapproachesandprocesses thatworkedforonephasebreakdownoncethescale-upreachesthenextphase.This book is designed to help you successfully navigate the phase changes
you’llfaceonthepathtoglobaldominance.Throughoutthisbook,wewillrefertothefivekeystagesofblitzscalingusing
themetaphor of a community. Since themost obvious, visible, and impactfulchangeinascale-upisthenumberofpeopleitemploys,we’lldefinethestagesbasedonthenumberofemployeesinthecompany,oritsorganizationalscale.
THEFIVESTAGESOFBLITZSCALING
Stage1(Family) 1–9employeesStage2(Tribe) 10sofemployees
Stage3(Village) 100sofemployeesStage4(City) 1000sofemployees
Stage5(Nation) 10000sofemployees
Each stage has critical differences when it comes to management andleadership.Whenyou’reheadofanuclearFamily,youhavecloserelationshipswith all of your Family members.When you’re the head of a whole Nation,
you’reresponsible for the livesofamultitudeofpeople,mostofwhomyou’llnevermeet. (Later in the book we’ll talk about how to optimize your peoplemanagementstrategyasyourcompanygrows.)It’simportanttorememberthatwhilethesepowersoftenprovideaclearand
consistent set of categories, real life is oftenmessier. For example, a start-upwith a tight-knit team might feel and act like a Family even if it has nearlytwentyemployees.Sothesedefinitionsaremeantsimplytoofferausefulsetofguidelines.We also recognize that the number of employees is only one of several
measuresofanorganization’sscale.Someoftheothermeasuresofscaleincludethenumberofusers(userscale),thenumberofcustomers(customerscale),andtotalannualrevenues(businessscale).Thesemeasuresusually,butdon’talways,move in lockstep. While it’s nearly impossible to achieve customer scale orbusinessscalewithoutorganizationalscale—customersrequirecustomerservicerepresentatives, and revenues typically require salespeople—it is possible toachieve user scale without organizational scale. Consider the example ofInstagram:whenthatcompanywasacquiredbyFacebookfor$1billion,ithadover one hundredmillion users but just thirteen employees and no significantrevenues.The fact that the phases don’t always move in lockstep is a feature of
blitzscaling, not a bug. As we’ll discuss, operational scalability is one of theprimary growth limiters that scale-ups need to address.When a business cangrow users, customers, and revenues faster than the number of employeeswithoutcollapsingundertheweightofitsowngrowth,thebusinesscanachievegreater profitability and keep growingwithout being as tightly constrained bythe need for financial or human capital. In contrast, when the number ofemployeesgrowsfasterthanusers,customers,andrevenues,it’samajorredflagthatcouldindicateissueswiththefundamentalbusinessmodel.Nevertheless, for the sake of simplicity, this book will typically define the
stage of a company by its organizational scale. A Family-stage companywillhaveonetonineemployees,aTribe-stagecompanywillhavetentoninety-nineemployees,andsoon.Whenexceptionsarise,we’llspecificallycallthemouttoavoidconfusion.
THETHREEKEYTECHNIQUESOFBLITZSCALING
Throughmuchstudyof,directaccessto,andconversationwiththeleadershipatcompanies such as Google, Amazon, and Facebook—and through my ownexperiencesasanentrepreneurandaninvestor—we’vebeenabletoidentifythethreekey techniques appliedby entrepreneurs and investors to builddominantcompanies. These basic principles do not depend on geography and can beadaptedtobuildgreatcompaniesinanyecosystem,albeitwithvaryingdegreesofdifficulty.
TECHNIQUE#1:BUSINESSMODELINNOVATION
Thefirsttechniqueofblitzscalingistodesignaninnovativebusinessmodelthatcan truly grow. This sounds like a Start-ups 101–level insight, but it’sastoundinghowmanyfoundersmissthiskeyelement.Amajormistakemadebymanystart-upsaroundtheworldisfocusingonthetechnology,thesoftware,theproduct,andthedesign,butneglecting toeverfigureout thebusiness.Andby“business”we simplymeanhow the companymakesmoneybyacquiring andserving its customers. In contrast, despite the popular “engineers are gods”narrativeprevalentinSiliconValley,thecompaniesandfoundersweuniversallyhail as geniuses aren’t just technology nerds—they’re almost always businessnerdstoo.AtGoogle,LarryPageandSergeyBrinbuiltgreatsearchalgorithms,but itwas their innovations to the search enginebusinessmodel—specifically,considering relevance and performancewhen displaying advertisements ratherthan simply renting space to the highest bidder—that drove their massivesuccess.As theworld has gone digital, businessmodel innovation has become even
more important. Somany technologies are available as services,which are ondemand and built to be integrated, that technology is no longer as strong adifferentiator, while figuring out the right combinations of services to bringtogetherintoabreakthroughproducthasbecomeamajordifferentiator.Mostoftoday’s successful companies are more like Tesla, which combines a set oftechnologiesthatalreadyexisted,ratherthanSpaceX,whichhadtopioneernewones.Businessmodelinnovationishowstart-upsareabletooutcompeteestablished
competitors who typically hold a host of advantages over any upstarts. As astart-up,DropboxcompeteswithgiantslikeMicrosoftandGoogle,whooughttohave major advantages in technology, finance, and market power. Dropbox
founderandCEODrewHoustonknowsthathiscompanycan’tsimplyrelyonbettertechnologyoroutexecutingthecompetition:“Ifyourplaybookisthesameasyourcompetitor’s,youareintrouble,becausechancesaretheyarejustgoingtorunyourplaybookwithalotmoreresources!”Drewhad to design a better businessmodel, inwhich the focus on sharing
filesmeans that the number of filesDropbox has to store (or in the past, payAmazon to store) increases far more slowly than the value created for thecustomerandthustherevenuesDropboxcancollectfromthosecustomers.Uberand Airbnb also built large businesses at incredible speed based on novelbusiness models rather than unprecedented new technologies. If technologicalinnovationalonewereenough,federalresearchlabswouldproduce$100billioncompaniesonaregularbasis.Spoileralert:theydon’t.This is not to say that technology innovation is unimportant. Technology
innovationisthemostcommontriggerforlaunchinganewmarketorupendinganexistingone.Uberwasn’tthefirstcompanytotrytoimprovetheexperienceof hailing a taxi.But prior to the technological innovation of the smartphone,complete with wireless Internet connection and GPS-enabled location-basedservices, Uber’s business model simply wouldn’t have worked. Theseinnovations reduced the friction for both driver and rider,makingUber’s coreUberXridesharingmodelamass-marketpossibilityforthefirsttime.Nor can companies afford to ignore technology innovation after they
successfullyblitzscaletheirwaytoCityorNationstage.Eachandeveryoneofthe technology companies worth over $100 billion has used technologyleadershiptoreinforceitscompetitiveadvantages.Amazonmayhavestartedasa simpleonline retailerwithnounique technology,but today its technologicalprowessincloudcomputing,automatedlogistics,andvoicerecognitionhelptomaintain its dominance. In fact, the megacompanies built by blitzscaling areoften the ones buying the technology innovators, much as Google boughtDeepMindandFacebookboughtOculus.Technology innovation is a key factor in retaining the gains produced by
businessmodel innovation.Afterall, ifonetechnologyinnovationcancreateanew market, another technology innovation can render it obsolete, seeminglyovernight. While Uber has achieved massive scale, the greatest threat to itsfuturedoesn’tcomeintheformofdirectcompetitorslikeDidiChuxing,thoughthese are formidable threats. The greatest threat to Uber’s business is thetechnologyinnovationofautonomousvehicles,whichcouldmakeobsoleteone
of Uber’s biggest competitive advantages—its carefully cultivated network ofdrivers—essentiallyovernight.The key is to combine new technologies with effective distribution to
potentialcustomers,ascalableandhigh-marginrevenuemodel,andanapproachthat allows you to serve those customers given your probable resourceconstraints.Ideally, you design your business model innovation before you start your
company.ThisiswhathappenedwhenIcofoundedLinkedIn.Thekeybusinessmodel innovations for LinkedIn, including the two-way nature of therelationships and filling professionals’ need for a business-oriented onlineidentity, didn’t just happen organically.Theywere the result ofmuch thoughtand reflection, and I drewon the experiences I hadwhen foundingSocialNet,one of the first online social networks, nearly a decade before the creation ofLinkedIn. But life isn’t always so neat. Many companies, even famous andsuccessfulones,havetodeveloptheirbusinessmodelinnovationaftertheyhavealreadycommencedoperations.PayPaldidn’thaveabusinessmodelwhen itbeganoperations (Iwasakey
member of the PayPal executive team). We were growing exponentially, at5 percent per day, andwewere losingmoney on every single transactionweprocessed.Thefunnythingisthatsomeofourcriticscalledusinsaneforpayingcustomers bonuses to refer their friends. Those referral bonuseswere actuallybrilliant, because their cost was so much lower than the standard cost ofacquiring new financial services customers via advertising. (We’ll discuss thepowerandimportanceofthiskindofviralmarketinglateron.)Theinsanity,infact,wasthatwewereallowingouruserstoacceptcreditcard
payments,stickingPayPalwiththecostofpaying3percentofeachtransactiontothecreditcardprocessors,whilechargingourusersnothing.IrememberoncetellingmyoldcollegefriendandPayPalcofounder/CEOPeterThiel,“Peter, ifyou and I were standing on the roof of our office and throwing stacks ofhundred-dollarbillsofftheedgeasfastasourarmscouldgo,westillwouldn’tbe losing money as quickly as we are right now.”We ended up solving theproblem by charging businesses to accept payments, much as the credit cardprocessors did, but funding those payments using automated clearinghouse(ACH)bank transactions,which cost a fraction of the charges associatedwiththecreditcardnetworks.Butifwehadwaiteduntilwehadsolvedthisproblembeforeblitzscaling,Isuspectwewouldn’thavebecomethemarketleader.
TECHNIQUE#2:STRATEGYINNOVATION
The most obvious element of blitzscaling is the pursuit of extreme growth,which,whencombinedwithaninnovativebusinessmodel,cangeneratemassivevalue and long-term competitive advantage. Many start-ups believe they arepursuingastrategyofextremegrowth,wheninfacttheyhavethegoalandthewishforextremegrowthbutnounderstandingofanactualstrategythatwillgetthemthere.Toachieveyourgoals,youhavetoknowwhatyouplantodoand,justasimportant,whatyouplannottodo.Also,growthdoesn’tcreatevalueinandofitself;forthat,ithastobepairedwithaworkingbusinessmodel.It’seasytoachieveextremecustomerandrevenuegrowthifyourcompanysells$20billsfor $1, but “we’ll make it up in volume” won’t allow you to build anysustainablevalue.Forsuccessfulblitzscaling,thecompetitiveadvantagecomesfromthegrowth
factorsbuiltintothebusinessmodel,suchasnetworkeffects,wherebythefirstcompany to achieve critical scale triggers a feedback loop that allows it todominate a winner-take-all or winner-take-most market and achieve a lastingfirst-scaler advantage. For example, Uber’s strategy of aggressive city-by-cityexpansion allows its customers to hail rides with fewer delays than itscompetitors.UberwantsyoutobeabletogetaridefasterwithUberthanwithanyone else. This attractsmore customers,which attractsmore drivers,whichincreasestheliquidityofthemarketplace,whichallowscustomerstohailrideseven more quickly, which attracts more customers, and so on. Early UberinvestorBillGurleylaidoutUber’sstrategyinhis2012blogpost“AllMarketsAreNotCreatedEqual.”
Asthecompanygrows,theyareabletofacilitatemorecarsontheroad, and along with their investment in route and loadoptimization, thisallowsforshorterandshorterpickuptimes.Theexperiencegetsbetterandbetterthelongertheyareinthemarket.
Blitzscaling goes beyond just a strategy of aggressive growth because itinvolves doing things that don’t make sense according to traditional businessthinking, such as prioritizing speed over efficiency despite an uncertainenvironment.Atthesametime,blitzscalingalsogoesbeyondjustrisktaking.Itmayberiskytobetthecompany,asWaltDisneydidwhenheborrowedagainsthisownlifeinsurancetobuildDisneyland,butit’snotblitzscaling.Blitzscaling
would have involved inefficiencies like paying construction crews to worktwenty-fourhoursadayinordertogetDisneylandopenafewmonthsearlier,orreducingticketprices90percenttogettoonemillionvisitorsfaster—knowingthatthoseonemillionvisitorswerenetworkedtotenmillionmore.Here isoneof the ruthlesspractices thathashelpedmakeSiliconValley so
successful:Investorswilllookatacompanythatisonanupwardtrajectorybutdoesn’tdisplaytheproverbialhockeystickofexponentialgrowthandconcludethat they need to either sell the business or take on additional risk thatmightincrease the chances of achieving exponential growth. Achieving 20 percentannual growth, which would delight Wall Street analysts covering any otherindustry, simply isn’t enough to transform a start-up into amultibillion-dollarcompany fast enough. SiliconValley venture capitalistswant entrepreneurs topursueexponentialgrowthevenifdoingsocostsmoremoneyandincreasesthechancesthat thebusinesscouldfail,resultinginabiggerloss.Droppingbeloweven40percentannualgrowthisawarningsignforinvestors.Thismindsetcanbedifficultforpeopletounderstand.“WhyshouldIriskit
allandpotentiallyblowupwhatisasuccessful,growingbusiness?”theymightrightfullyask.Theansweristhatblitzscalingbusinessestendtoplayinwinner-take-mostorwinner-take-allmarkets.Thegreaterriskforasuccessful,growingbusiness is to move too slowly and allow its competitors to win marketleadershipandfirst-scaleradvantage.Nokia is a great example of the cost of caution. In 2007, Nokia was the
world’s largest and most successful maker of mobile phones, with a marketcapitalizationofjustunder$99billion.ThenAppleandSamsungcameblazinginto the market. In 2013, Nokia sold its money-losing handset operations toMicrosoftfor$7billion,andin2016MicrosoftsolditsfeaturephoneassetsandtheNokiahandsetbrand toFoxconnandHMDfor just$350million.That’sadrop in value for Nokia’s mobile phone business from somewhere in theneighborhoodof$99billionto$350millioninlessthanadecade—adeclineofover99percent.At the time, Nokia’s decisions may have seemed to make sense. Nokia
actually continued growing even after the launch of the iPhone andGoogle’sAndroid operating system.Nokia hit its peak in termsof unit volumewhen itshipped104millionphones in2010.ButNokia’ssalesdeclinedafter that,andwere surpassed byAndroid in 2011 and iPhone in 2012.By the timeNokia’smanagementrealizedtheexistentialthreatfacingthem,itwastoolate;eventhe
desperationplayofaligningthemselveswithMicrosoftasitsexclusiveWindowsPhonepartnercouldn’treversethedecline.Becauseblitzscalingoftenrequiresspendingsignificantamountsofcapitalin
waysthattraditionalbusinesswisdomwouldconsider“wasteful,”implementinga financial strategy that supports this aggressive spending is a critical part ofblitzscaling.Forexample,Uberoftenusesheavysubsidiesonbothsidesofthemarketplacewhenitlaunchesinanewcity,loweringfarestoattractridersandboosting payments to attract drivers. By paying out more than it takes in onthose early trips, Uber is able to reach critical scale faster than a moreconservative competitor.Given thewinner-take-most nature of the ridesharingmarket, that “wasteful” spending has helpedUber achieve a dominantmarketpositioninthecitiesinwhichitoperates.Ofcourse,thatstrategyisn’tpossiblewithout the ability to raisemassive amounts of capital on favorable terms. InUber’scase,ithasbeenabletoraisenearly$9billionbetweenitsfoundingandthewritingofthisbook.Atsomepoint,Uberwillhavetodemonstratetheabilitytosignificantlyimproveitsuniteconomics,oritsinvestorswillgetverygrumpy.This concern helps explain Uber’s significant investments in autonomousvehicletechnology,whichcouldeliminateitsbiggestexpense—driverpayments—inonefellswoop.Thewillingnesstotakeontherisksofblitzscalingisoneofthemajorreasons
whySiliconValley has produced such a disproportionate share of blockbustercompaniesincomparisontoothergeographies.Tobefair,ithasalsoproducedadisproportionate share of financial disasters—hence the word “risk” whentalkingaboutblitzscaling.ButastheriseofjuggernautslikeAlibabaandSpotifyillustrates,blitzscalingisalsostartingtotakeoffaroundtheworld.
TECHNIQUE#3:MANAGEMENTINNOVATION
Thefinal techniquerequiredforblitzscalingismanagementinnovation.Thisisnecessary because of the extreme strains placed on the organization and itsemployeesbyhypergrowth.Iamfondofpointingouttoentrepreneursandexecutivesthat“intheory,you
don’tneedpractice.”WhatImeanisthatnomatterhowbrilliantyourbusinessmodelandgrowth
strategy, you won’t be able to build a real-world (i.e., non-theoretical)blockbuster companywithout a lot of practice. But that problem ismagnified
whenyou’retryingtoblitzscale.The kind of growth involved in blitzscaling typically means major human
resources challenges. Tripling the number of employees each year isn’tuncommon for a blitzscaling company. This requires a radically differentapproachtomanagementthanthatofatypicalgrowthcompany,whichwouldbehappytogrow15percentperyearandcantaketimefindingafewperfecthiresandobsessingaboutcorporateculture.Aswewilldiscussinmoredetaillaterinthe book, companies that blitzscale have to rapidly navigate a set of keytransitions as their organizations grow, and have to embrace counterintuitiverules like hiring “good enough” people, launching flawed and imperfectproducts,lettingfiresburn,andignoringangrycustomers.Overthecourseofthisbook,we’llseehowbusinessmodel,growthstrategy,
andmanagement innovationwork together to form the high-risk, high-rewardprocessofblitzscaling.
PARTII
BusinessModelInnovation
Ofthethreecoretechniquesofblitzscaling,thefirstandmostfoundationalistodesignaninnovativebusinessmodelcapableofexponentialgrowth.The story of entrepreneurship in the Internet era is a story of this kind of
businessmodelinnovation.Think back to the dot-com era, which stretched roughly from the IPO of
Netscapein1995untiltheNASDAQbegantocrashin2000.Duringthisperiod,enormousnumbersofstart-upsandprettymucheveryestablishedcompanytriedtobuildgreatInternetbusinesses,yetnearlyallofthemfailed.Theproblemwas,mostofthemsimplytriedtocutandpasteexistingbusinessmodelsontothenewonlinemedium.Youcan’t transplantaheartfromonespecies intoanotherandexpectittothrive.If youhad asked stockmarket analysts in1995which companieswerebest
positionedtodominatetheInternet,mostwouldhavepointedtoexistinggiantslikeMicrosoftandTimeWarner,whichinvestedmillionsinInternetbusinesseslikeMSN and Pathfinder.Otherswould havementioned “pure play” dot-comstart-upslikeeToys,whichcombinedprovenbusinessmodelslikethe“categorykiller”storewiththenewonlinemedium.Yet when the wreckage of the dot-com crash cleared, the most successful
companies still charging full steam ahead were the few start-ups that weredesigned around totally new business models, such as Amazon, eBay, andGoogle.Walmart should have dominated online retail, yet Amazon emerged and
practically wrote the bible for e-commerce, including consumer reviews,
shopping carts, and free shipping. Newspapers and phone book companiesshould have been able to transfer their information businesses to the onlineworld,butYahoo!andthenGooglesteppeduptotheplate.Theybuiltthesearchengines that indexed the world’s information, and Google developed thebusinessmodel thatmade it worthmore than all traditionalmedia companiescombined.In contrast, and much to their misfortune, start-ups that relied purely on
technologyinnovationwithoutanyrealbusinessmodelinnovationlargelywentbust.CompanieslikeeToysthattriedto“Amazon”variousmarkets,butwithoutAmazon’s front-and back-office innovations, crashed and burned once thefinancial markets began to demand profits rather than just expensive revenuegrowth. Even Netscape, whose Netscape Navigator mainstreamed Webbrowsing,andwhoseIPOkickedoffthedot-comboom,wasforcedtosellitselfofftoAOL.NetscapeengineersinventedJavaScript,SSL,andallkindsofcooltechnology for the Internet that are stillused today,butNetscapeaccepted thestatus quo when it came to using tried-and-true business models rather thandeveloping new ones that were enabled by its own technology innovation.Unfortunately forNetscape, its competitorMicrosoft already understood thosebusinessmodels all toowell andknewexactlyhow touse its economicmightand resources to pull their levers. In the first “browser war,” Microsoftpreinstalled its Internet Explorer on all new Windows computers, then gaveaway its Web server software, Internet Information Server (IIS), whicheffectivelydestroyedNetscape’sbusinessmodel.Could Netscape have succeeded with a different strategy? We believe so.
Consider that one of theways thatNetscapemonetized itsNavigator browserwastosellthesponsorshipofitsNetSearchbuttontotheExcitesearchenginefor $5 million. Netscape believed that the browser itself was the key, whilesearch was simply a sideline. It was left to two pairs of Stanford graduatestudents,JerryYangandDavidFilo(Yahoo!)andLarryPageandSergeyBrin(Google),toprovethatsearchwasamuchbiggerbusiness.Google’sinnovativemodel of selling text ads next to search results via an automatedmarketplaceallowed it to build a franchise so dominant that it later withstood a series offrontalassaultsbyMicrosoft,includingamarketingprograminwhichMicrosoftessentiallypaidpeopletouseitsBingsearchengine.The same story has been repeated in multiple waves since. Facebook and
LinkedIndominatesocialnetworkseventhoughAOL,Microsoft(Hotmail),andYahoo! (Yahoo!Mail) controlledmost consumer online identitieswhen those
socialnetworksfirstemerged.AlibababeateBayinChina.Uberoutflankedthetaxi companies.Airbnbhasmore room listings thananyhotel company in theworld.These success stories are technology companies, sure. But as we’ve seen,
technological innovation alone is insufficient—even when its impact on thefuture is huge. Services like Craigslist, Wikipedia, and IMDb (the InternetMovieDatabase)wereearly,influentialInternetinnovators,buttheystillneverbecamemassively(financially)valuableontheirown.Therealvaluecreationcomeswheninnovativetechnologyenablesinnovative
products and services with innovative business models. Even though thebusinessmodelsofGoogle,Alibaba,andFacebookmightseemobvious—eveninevitable—after the fact, they weren’t widely appreciated at the time theylaunched.Howmanypeoplein1999wouldhaverealizedthatrunningtinytextadsnexttotheequivalentofanelectroniccardcatalogwouldleadtotheworld’smostvaluablesoftwarecompany?OrthatsettingupanonlineshoppingmallforChina’semergingmiddleclasswouldleadtoa$100billionbusiness?Whichofyouin2004wouldhavepredictedthatlettingpeopleseewhattheirfriendsaretalkingaboutbystaringatatinyscreenonahandheldcomputerwouldbecomethedominantformofmedia?Greatcompaniesandgreatbusinessesoftenseemtobebadideaswhentheyfirstappearbecausebusinessmodelinnovations—bytheir very definition—can’t point to a proven business model to demonstratewhythey’llwork.Toreallyunderstandwhythesebusinessmodelssucceed,weneedtoclearly
definewhatwemeanby“businessmodel”inthefirstplace.Partoftheproblemis that the term can be interpreted in so many different ways. The greatmanagement thinker Peter Drucker wrote that business models are essentiallytheories composed of assumptions about the business, which circumstancesmight require to change over time. Harvard Business School professor andauthorClayChristensenbelieves that youneed to focuson the concept of the“job-to-be-done”; that is,whenacustomerbuysaproduct,she is“hiring”it todo a particular job. Then there’s Brian Chesky of Airbnb, who said simply,“Build a product people love.Hire amazing people.What else is there to do?Everythingelseisfakework.”AsAndreaOvansaptlyput it inherJanuary2015HarvardBusinessReview
article,“WhatIsaBusinessModel?”,it’senoughtomakeyourheadswim!Forthe purposes of this book, we’ll focus on the basic definition: a company’s
business model describes how it generates financial returns by producing,selling,andsupportingitsproducts.What sets companies likeAmazon,Google, andFacebook apart, even from
othersuccessfulhigh-techcompanies,isthattheyhaveconsistentlybeenabletodesign and execute business models with characteristics that allow them toquicklyachievemassivescaleandsustainablecompetitiveadvantage.Ofcourse,there isn’t a single perfect businessmodel thatworks for every company, andtryingtofindoneisawasteoftime.Butmostgreatbusinessmodelshavecertaincharacteristics in common. If youwant to find your best businessmodel, youshouldtrytodesignonethatmaximizesfourkeygrowthfactorsandminimizestwokeygrowthlimiters.
DESIGNINGTOMAXIMIZEGROWTH:THEFOURGROWTHFACTORS
GROWTHFACTOR#1:MARKETSIZE
Themostbasicgrowthfactortoconsiderforyourbusinessmodelismarketsize.Thisfocusonmarketsizemaysoundobvious,andit’srightoutofPitchDeck101forstart-ups,butifyouwanttobuildamassivecompany,youneedtobeginwiththebasicsandeliminateideasthatservetoosmallofamarket.Abigmarkethasbothalargenumberofpotentialcustomersandavarietyof
efficient channels for reaching those customers.That last point is important; amarket consisting of “everyone in the world” might seem large, but it isn’treachableinanyefficientway.We’lldiscussthisingreaterdepthwhenwelookatdistributionasakeygrowthfactor.It’snot easy to judge the sizeof amarket, orwhatpitchdecks andventure
capitalistsoften refer toasTAM(totalavailablemarket).PredictingTAMandhow it will grow in the future is one of the main sources of uncertainty inblitzscaling. But predicting it correctly and investing accordinglywhen othersarestillparalyzedbyfearisalsooneofthemainopportunitiesforunexpectedlyhighreturns,aswe’llseeinthecasesofAirbnbandUber.Ideally, themarket itself isalsogrowingquickly,whichcanmakeasmaller
marketattractiveandalargemarketirresistible.InSiliconValley,thecompetitionforventurecapitalexertsastrongpressure
on entrepreneurs to focus on ideas that are going after big markets. Venture
capital firmsmight raise hundreds ofmillions or evenbillions of dollars fromtheirinvestors—limitedpartnerslikepensionfundsanduniversityendowments—whoareseekingabove-marketreturnstocompensatethemfortakingachanceonprivatelyheldcompanies rather than simply investing in theCoca-Colasoftheworld.Todelivertheseabove-marketreturns,venturecapitalfundsneedtoatleasttripletheirinvestors’money.A$100millionventurecapitalfundwouldneed to return$300millionover the typical seven-to ten-year lifeofa fund toachieveanabove-marketinternalrateofreturnof15to22percent.A$1billionfund would need to return $3 billion. Since most venture capital investmentseither lose money or barely break even, the only realistic way that venturecapitalists can achieve these aggressive goals is to rely on a small number ofincredibly successful investments. For example, Benchmark Capital invested$6.7millionineBayin1997.Lessthantwoyearslater,eBaywentpublic,andBenchmark’s stake was worth $5 billion, which is a 745 times return. Thespecific fund that made that investment, Benchmark Capital Partners I, took$85millionfrominvestorsandreturned$7.8billion,fora92timesreturn.(Theinitial investors in Facebook did even better, but were individuals rather thanfirms.)Given the desire for home runs like eBay, most venture capitalists filter
investment opportunities based on market size. If a company can’t achieve“venturescale”(generally,amarketofat least$1billioninannualsales), thenmostVCswon’tinvest,evenifitisagoodbusiness.Itsimplyisn’tlargeenoughto help them achieve their goal of returning more than three times theirinvestors’money.WhenBrianCheskywaspitchingventurecapitaliststoinvestinAirbnb,one
ofthepeopleheconsultedwastheentrepreneurandinvestorSamAltman,wholaterbecamethepresidentoftheYCombinatorstart-upaccelerator.AltmansawChesky’spitchdeckandtoldhimitwasperfect,exceptthatheneededtochangethemarket-sizeslidefromamodest$30millionto$30billion.“InvestorswantB’s,baby,”AltmantoldChesky.Ofcourse,Altmanwasn’ttellingCheskytolie;rather,hearguedthatiftheAirbnbteamtrulybelievedintheirownassumptions,$30millionwasagrossunderestimate,andtheyshoulduseanumberthatwastrue to theirconvictions.As it turnsout,Airbnb’smarketwas indeedcloser to$30billion.Whenevaluatingmarketsize,it’salsocriticaltotrytoaccountforhowlower
costs and product improvements can expand markets by appealing to newcustomers, in addition to seizingmarket share from existing players. In 2014,
AswathDamodaran,aprofessoroffinanceatNYU’sSternSchoolofBusiness,estimatedthatUberwasprobablyworthroughly$6billion,basedonitsabilityto ultimately win 10 percent of the global taxi market of $100 billion, or$10 billion. According to Uber’s own projections, in 2016 the companyprocessed over $26 billion in payments. It’s safe to say that the $10 billionmarketwasaseriousunderestimate,as theeaseofuseand lowercostofUberanditscompetitorsexpandedthemarketfortransportation-as-a-service.AsAaronLevie,thefounderoftheonlinefilestoragecompanyBoxnotedin
a tweet in 2014, “Sizing the market for a disruptor based on an incumbent’smarketislikesizingacarindustryoffhowmanyhorsestherewerein1910.”The other factor that can lead to underestimating amarket is neglecting to
account for expanding into additional markets. Amazon began as AmazonBooks, the “Earth’s Biggest Bookstore.” But Jeff Bezos always intended forbooksellingtoserveasabeachheadfromwhichAmazoncouldexpandoutwardto encompass his massive vision of “the everything store.” Today, Amazondominates the bookselling industry, but thanks to relentlessmarket expansion,booksalesrepresentlessthan7percentofAmazon’stotalsales.The same effect can be seen in the financial results of Apple. In the first
quarter of 2017, Apple generated $7.2 billion from the sale of personalcomputers, a category the company pioneered and once dominated. That’s agreat number to be sure, but, over that same financial quarter, Apple’s totalrevenuewasawhopping$78.4billion,whichmeantthatApple’soriginalmarketaccountedforlessthan10percentofitstotalsales.My Greylock colleague Jerry Chen, who helped Diane Greene scale
VMware’s virtualization software into a massive business, likes to point out,“Everybillion-dollarbusinessstartedasaten-million-dollarbusiness.”Butwhetheryouarecreatinganewmarket,expandinganexistingmarket,or
relyingonadjacentmarketstogettothose“B’s”thatinvestorswant(baby),youneedtohaveaplausiblepathtogetfromheretothere.Thisleadsustooneofmyfavoritegrowthfactorstodiscusswithentrepreneurs:distribution.
GROWTHFACTOR#2:DISTRIBUTION
Thesecondgrowthfactorneededforastrong,scalablebusinessisdistribution.ManypeopleinSiliconValleyliketofocusonbuildingproductsthatare,inthe
famous words of the late Steve Jobs, “insanely great.” Great products arecertainlyapositive—we’lldiscussthelackofproductqualityasagrowthlimiterlater on—but the cold and unromantic fact is that a good product with greatdistributionwillalmostalwaysbeatagreatproductwithpoordistribution.Dropbox is a companywith agreatproduct,but it succeededbecauseof its
greatdistribution.InaninterviewforReid’sMastersofScalepodcast,founderandCEODrewHoustonsaid thathebelieves that toomanystart-upsoverlooktheimportanceofdistribution:
Most of the orthodoxy inSiliconValley is about building a goodproduct.Ithinkthat’sbecausemostcompaniesintheValleydon’tsurvive beyond the building-the-product phase. You have to begood at building a product, then you have to be just as good atgetting users, then you have to be just as good at building abusinessmodel.Ifyou’remissinganyofthelinksinthechain,thewholechainisbroken.
The challenge of distribution has become even greater in the “mobile first”era.Unlike theWeb,where search engine optimization and e-mail linkswerebroadlyapplicableandsuccessfuldistributionchannels,mobileappstoresofferlittleopportunityforserendipitousproductdiscovery.Whenyougo toApple’sorGoogle’sappstore,you’researchingforaspecificproduct.Fewpeopleinstallappsjustforthehellofit.Asaresult,thebusinessmodelinnovatorswhohavesucceeded(e.g.,Instagram,WhatsApp,Snap)havehadtofindcreativewaystogetbroaddistributionfortheirproduct—withoutspendingalotofmoney.Thesedistribution techniques fall into two general categories: leveraging existingnetworksandvirality.
A)LeveragingExistingNetworks
Newcompanies rarely have the reachor resources to simplypourmoney intoadvertising campaigns. Instead, they have to find creative ways to tap intoexistingnetworkstodistributetheirproducts.When I was at PayPal, one of the major vehicles for distribution of our
paymentservicewassettlingpurchasesoneBay.Atthetime,eBaywasalreadyoneofthelargestplayersine-commerce,andbythebeginningof2000already
had ten million registered users. We tapped into this user base by buildingsoftware that made it extremely easy for eBay sellers to automatically add a“PaywithPayPal”buttontoalloftheireBaylistings.Theamazingthingisthatcustomers did so even though eBay had its own rival payments service,Billpoint!But sellerswere required to addBillpointmanually to each of theirlistings;PayPaldiditforthem.Manyyearslater,Airbnbwasabletoperformasimilarfeatbyleveragingthe
onlineclassifiedserviceCraigslist.BasedonasuggestionfromYCombinator’sMichaelSeibel,Airbnbbuiltasystemthatallowedandencourageditshosts tocross-post their listings to the much-larger Craigslist. Hosts were told,“Reposting your listing from Airbnb to Craigslist increases your earnings by$500 a month on average,” and were allowed to do so by clicking a singlebutton. This took serious technology skills—unlikemany platforms,Craigslistdoesn’t have an application programming interface (API) that allows othersoftwaretointeractwithit—butitwastechnologyinnovationforthepurposesofdistribution innovation, not product innovation. “It was a kind of a novelapproach,” Airbnb founder Nathan Blecharczyk said of the integration. “Noothersitehadthatslickanintegration.Itwasquitesuccessfulforus.”Leveraging an existing network can have downsides, of course. What the
existingnetworkgives(orunknowinglyallowstobetaken),theexistingnetworkcan also take away.Zynga, the leading social games company, achievedgreatsuccessleveragingFacebookfordistribution,buthadtodramaticallyreengineerits distributionmodel after Facebook decided to stop allowing people playingZyngagamestoposttheirprogresstotheirFacebookfriends.(Disclosure:Iamamember of Zynga’s board of directors.) Zynga founder Mark Pincus wasfarsightedenoughtobuildastrongenoughfranchisetosurvivethechange.In contrast, so-called content farms like Demand Media that leveraged
Google’s search platform to generate website traffic and advertising revenuesnever recovered afterGoogle tuned its algorithms to deprioritize content fromwhatitcalled“junk”websites.Despitethesedangers,leveragingexistingnetworkscanbeacriticalpartofa
businessmodel,especiallyifthesenetworkscanprovidea“boosterrocket”thatislatersupplementedwithviralityornetworkeffects.
B)Virality
“Viral” distribution occurswhen the users of a product bringmore users, andthose users bring additional users, and so on, much like an infectious virusspreadsfromhost tohost.Viralitycaneitherbeorganic—occurringduring thecourseofnormalusageoftheproduct—orincentivizedbysomekindofreward.AfterlaunchingLinkedIn,theteamandIdevotedsignificanttimeandenergy
tofiguringouthowtoimproveorganicvirality;thatis,howtomakeiteasierforexistingusers to invite friends touse the service.Onewaywedid thiswas torefinewhathavebecomesomeofthestandardtoolsofvirality,suchasaddressbook importers. For example, we built software that allowed LinkedIn toconnect to our users’ Outlook contacts, whichmade it very easy for them toinvitetheirmostimportantconnections.Butequallyimportantwasanunanticipatedsourceofvirality.Asitturnedout,
userswanted touse theirLinkedInpagesas theirprimaryprofessional identityontheInternet.Havingapagelikethistopointothersto—withallthedetailsoftheir professional life together in one place—generated value not only for theuser,butforthepeopleviewingthepage,anditmadeviewersrealizethattheyshouldgettheirownLinkedInprofile.Asaresult,weaddedpublicprofilesasasystematictooltoboostboththemembervaluepropositionandourviralgrowthrate.At PayPal, we combined organic and incentivized virality. The payment
productwasinherentlyviral;ifsomeonee-mailedyoumoneyusingPayPal,youhadtosetupanaccounttogetpaid.Butweenhancedthisorganicviralitywithmonetary incentives. Ifyou referreda friend toPayPal,yougot$10,andyourfriend got $10. This combination of organic and incentivized virality allowedPayPal to grow 7 to 10 percent per day. As the PayPal network grew, wereducedtheincentivesto$5and$5,thenfinallyeliminatedthemaltogether.Incentives don’t have to bemonetary; like PayPal, Dropbox used a similar
combination of organic virality (as users share files with nonusers) andincentivizedvirality(Basicaccountholdersget500MBofextrastorageperuserthey refer; Pro account holders get 1 GB) to grow. Even though DropboxinvestedinpartnershipswithleadingPCmakerslikeDell,DrewHoustoncreditsvirality with driving the company’s rapid growth, helping it double its onehundred thousandusers at launch to twohundred thousandusers just tendayslater,thenskyrockettoonemillionusersjustsevenmonthsafterthat.If your distribution strategy focuses on virality, you also have to focus on
retention.Bringingnewusersinthroughthefrontdoordoesn’thelpyougrowif
they immediately turn around and leave. According to Houston, Dropboxdiscovered this truth the hard way, when activation rates revealed that only40percentofthepeoplesigningupwereactuallyputtingfilesintheirDropboxand linking them to their computers. In an interview formyMasters of Scalepodcast, Drew described a scene reminiscent of the television show SiliconValley(butwithahappierending):
Whatwe did iswewent onCraigslist and offered $40 to anyonewho’dcomeinforhalfanhour—apoorman’susabilitytest.We’relike,“Allright,sitdown.ThisisaninvitationtoDropboxinyoure-mail.Gofromheretosharingafilewiththise-mailaddress.”Zeroofthefivepeoplewetestedsucceeded.Zeroofthefiveevencameclose.Thiswasjuststunning.We’relike,“OhmyGod,thisis theworstproductevercreated.”Sowemadealistoflikeeightythingsin this Excel spreadsheet, then just sanded down all these roughedgesintheexperience,andwatchedouractivationrateclimb.
Viralityalmostalwaysrequiresaproductthatiseitherfreeorfreemium(i.e.,freeuptoacertainpoint,afterwhichtheuserhastopaytoupgrade—Dropbox,forexample,offers2GBoffreestorage).Wecan’trecallasingleinstanceofacompany that grew to a massive scale by leveraging the virality of a paidproduct.One of the most powerful distribution innovations is to combine both
strategies.Facebookwasable todo thisbyharnessing theorganicviralityofasocial network (where users invite other users to join them) and leveragingexisting networks centered around campuses by rolling out the product on acollege-by-college basis. We’ll discuss Facebook’s rollout strategy in greaterdepthwhenweconsidernetworkeffects.
GROWTHFACTOR#3:HIGHGROSSMARGINS
Oneofthekeygrowthfactorsthatentrepreneursoftenoverlookisthepowerofhigh gross margins. Gross margins, which represent sales minus the cost ofgoods sold, are probably the best measure of long-term unit economics. Thehigherthegrossmargin,themorevaluableeachdollarofsalesistothecompanybecause it means that for each dollar of sales, the company has more cash
available to fundgrowth andexpansion.Manyhigh-techbusinesseshavehighgrossmarginsbydefault,whichiswhythisfactorisoftenoverlooked.Softwarebusinesseshavehighgrossmarginsbecause thecostofduplicatingsoftware isessentiallyzero.Software-as-a-service (SaaS)businesseshaveaslightlyhighercostofgoods soldbecause theyneed tooperatea service,but thanks tocloudproviderslikeAmazon,thiscostisbecomingsmallerallthetime.Incontrast,“oldeconomy”businessesoftenhavelowgrossmargins.Growing
wheatisalow-marginbusiness,asissellinggoodsinastoreorservingfoodinarestaurant.OneofthemostamazingthingsaboutAmazon’ssuccessisthatithasbeen able to build amassive business based on retailing,which is generally alow-margin industry.AndevenAmazonnowreliesheavilyon itshigh-marginSaaS business, Amazon Web Services (AWS). In 2016, AWS accounted for150percentofAmazon’soperatingincome,whichmeansthattheretailbusinessactuallylostmoney.Most of the valuable companieswe’re focusing on in this book have gross
marginsofover60,70,oreven80percent.In2016,Googlehadagrossincomeof $54.6 billion on sales of $89.7 billion, for a gross margin of 61 percent.Facebook’sgrossincomewas$23.9billiononsalesof$27.6billion,foragrossmargin of 87 percent. In 2015, LinkedIn’s gross margin was 86 percent. Aswe’ve already discussed,Amazon is the outlier,with a 2016 gross income of$47.7billionon$136billioninsales,foragrossmarginof35percent.YetevenAmazon’s grossmargins are greater than those of a “highmargin” traditionalcompany like General Electric, which in 2016 had a gross income of$32.2billiononsalesof$119.7billion,foragrossmarginof27percent.Highgrossmarginsareapowerfulgrowthfactorbecause,asnotedbelow,not
all revenue is created equal. The key insight here is that even though grossmarginsmatteragreatdeal to theseller, theyare irrelevant to thebuyer.Howoften do you consider the grossmargin involvedwhen youmake a purchase?Wouldyou ever chooseBurgerKingoverMcDonald’s becauseWhoppers arelowermarginthanBigMacs?Typically,youfocussolelyonthecosttoyou,andtheperceivedbenefitsof thepurchase.Thismeansthat it’snotnecessarilyanyeasiertosellalow-marginproductthanahigh-marginproduct.Ifpossiblethen,acompanyshoulddesignahigh-gross-marginbusinessmodel.Second, high-gross-margin businesses are attractive to investors, who will
oftenpayapremiumfor thecash-generatingpowerof suchabusiness.As theprominentinvestorBillGurleywroteinhis2011blogpost,“AllRevenueIsNot
CreatedEqual,”“Investorslovecompanieswhere,allthingsbeingequal,higherrevenuescreatehigherprofitmargins.Sellingmorecopiesofthesamepieceofsoftware (with zero incremental costs) is a business that scales nicely.”Appealingtoinvestorsmakesiteasiertoraiselargeramountsofmoneyathighervaluationswhen the company is privately held (we’ll delve into the details ofwhy this is so important later on), and lowers the cost of capital when thecompanyispubliclytraded.Thisaccesstocapitalisakeyfactorinbeingabletofinancelightning-fastgrowth.It’s important to note the difference between potential gross margin and
realized gross margin. Many blitzscalers, such as Amazon or the Chinesehardware makers Huawei and Xiaomi, deliberately price their products tomaximize market share rather than gross margins. As Jeff Bezos is fond ofsaying,“Yourmarginismyopportunity.”Xiaomiexplicitlytargetsanetmarginof1to3percent,apracticeitcreditsCostcoforinspiring.Allotherfactorsbeingequal, investors almost always place amuch higher value on companies withhigherpotentialgrossmarginsthancompaniesthathavealreadymaximizedtheirrealizedgrossmargins.Finally,mostofacompany’soperationalchallengesscalebasedonrevenues
or unit sales volume, not grossmargin. If you have amillion customers whogenerate$100millionperyearinsales,thecosttoservethosecustomersdoesn’tchangewhetheryourgrossmarginis10percentor80percent;youstillneedtohireenoughpeople to respond to theirsupport requests.But it’sa loteasier toafford good customer support when you have $80million in grossmargin tospendratherthan$10million.Conversely, it’s a lot easier to sell and service 125,000 customers who
generate$12.5millionperyearinsalesand$10millioningrossmarginthanitisto have to sell and service amillion customerswho generate $100million insales to achieve that same $10million in grossmargin. That’s eight times asmanycustomersandeighttimestherevenues,whichmeanseighttimesasmanysalespeople,customerservicerepresentatives,accountants,andsoon.Designingahigh-gross-marginbusinessmodelmakesyourchancesofsuccess
greaterandtherewardsofsuccessevengreater.Aswe’llseeinalatersection,high grossmargins have helped even nontech businesses, such as the SpanishclothingretailerZara,growintoglobalgiants.
GROWTHFACTOR#4:NETWORKEFFECTS
Marketsize,distribution,andgrossmarginsare importantfactors ingrowingacompany,butthefinalgrowthfactorplaysthekeyroleinsustainingthatgrowthlongenoughtobuildamassivelyvaluableandlastingfranchise.Whilethepasttwentyyearshavedrivenimprovementsinthefirstthreegrowthfactors,therisein Internet usage around theworld has pushed network effects to levels neverbeforeseeninoureconomy.Theincreasingimportanceofnetworkeffectsisoneofthemainreasonsthat
technologyhasbecomeamoredominantpartoftheeconomy.At the end of 1996, the five most valuable companies in the world were
General Electric, Royal Dutch Shell, the Coca-Cola Company, NTT (NipponTelegraph and Telephone), and ExxonMobil—traditional industrial andconsumercompanies that reliedonmassiveeconomiesofscaleanddecadesofbrandingtodrivetheirvalue.Justtwenty-oneyearslater,inthefourthquarterof2017, the list looked very different: Apple, Google, Microsoft, Amazon, andFacebook.That’s a remarkable shift. Indeed,whileApple andMicrosoftwerealreadyprominentcompaniesat theendof1996,Amazonwasstillaprivatelyheldstart-up,LarryPageandSergeyBrinwerestillapairofgraduatestudentsatStanford who were two years away from founding Google, and MarkZuckerbergwasstilllookingforwardtohisbarmitzvah.Sowhathappened?TheNetworkedAgehappened,that’swhat.Technology now connects all of us in ways that were unthinkable to our
ancestors.Overtwobillionpeoplenowcarrysmartphones(manyofthemmadeby Apple, or using Google’s Android operating system) that keep themconstantly connected to the global network of everything. At any time, thosepeoplecanfindalmostany information in theworld (Google),buyalmostanyproductintheworld(Amazon/Alibaba),orcommunicatewithalmostanyotherhumanintheworld(Facebook/WhatsApp/Instagram/WeChat).Inthishighlyconnectedworld,morecompaniesthaneverareabletotapinto
networkeffectstogenerateoutsizegrowthandprofits.We’llusethesimplelayman’sdefinitionofnetworkeffectsinthisbook:
A product or service is subject to positive network effects whenincreasedusageby anyuser increases the value of the product orserviceforotherusers.
Economistsrefertotheseeffectsas“demand-sideeconomiesofscale”or,moregenerally,“positiveexternalities.”Themagicofnetwork effects is that theygenerate a positive feedback loop
that results in superlinear growth and value creation. This superlinear effectmakesitverydifficultforanynodeinthenetworktoswitchfromanincumbenttoanalternative(“customerlock-in”),sinceitisalmostimpossibleforanynewentranttomatchthevalueofpluggingintotheexistingnetwork.(Nodesinthesenetworksaretypicallycustomersorusers,asinthecanonicalexampleofthefaxmachine,orthemorerecentexampleofFacebook,butcanalsobedataelementsorotherfundamentalassetsvaluableinabusiness.)Theresultingphenomenonof“increasingreturnstoscale”oftenresultsinan
ultimateequilibriuminwhichasingleproductorcompanydominatesthemarketandcollects themajorityof its industry’sprofits.So it’snosurprise thatsmartentrepreneurs strive to create (and smart investors want to invest in) thesenetworkeffectsstart-ups.Severalgenerationsofstart-upshavetappedthesedynamicstobuilddominant
positions, from eBay to Facebook to Airbnb. To accomplish these goals, it’scritical to develop a rigorous understandingof hownetwork effectswork.MyGreylock colleague SimonRothman is one of theworld’s premier experts onnetwork effects from building eBay’s $14 billion automotive marketplace.Simonwarns,“Alotofpeopletrytoboltonnetworkeffectsbydoingthingslikeaddingaprofile.‘Marketplaceshaveprofiles,’theyreason,‘soifIaddprofiles,I’llbeaddingnetworkeffects.’ ”Yettherealityofbuildingnetworkeffectsisabit more complicated. Rather than simply imitate specific features, the bestblitzscalers study the different types of network effects and design them intotheirbusinessmodels.
FiveCategoriesofNetworkEffects
On his industrial organization of information technology website, the NYUprofessor Arun Sundararajan classifies network effects into five broadcategories:
1. Direct Network Effects: Increases in usage lead to direct increases invalue.(Examples:Facebook,messagingappslikeWeChatandWhatsApp)
2. Indirect Network Effects: Increases in usage encourage consumption ofcomplementary goods, which increases the value of the original product.(Example: Adoption of an operating system such asMicrosoftWindows,iOS, or Android encourages third-party software developers to buildapplications,increasingthevalueoftheplatform.)
3. Two-Sided Network Effects: Increases in usage by one set of usersincreasesthevaluetoadifferentsetofcomplementaryusers,andviceversa.(Example:MarketplacessuchaseBay,Uber,andAirbnb)
4. Local Network Effects: Increases in usage by a small subset of usersincreases the value for a connected user. (Example: Back in the days ofmetered calls, certain wireless carriers allowed subscribers to specify alimitednumberof“favorites”whosecallsdidn’tcountagainstthemonthlyallotmentofcallminutes.)
5. Compatibility and Standards: The use of one technology productencouragestheuseofcompatibleproducts.(Example:withintheMicrosoftOfficesuite,Word’sdominancemeantthatitsdocumentfileformatbecamethestandard;thishasallowedittodestroycompetitorslikeWordPerfectandfendoffopen-sourcesolutionslikeOpenDocument.)
Anyof thesedifferentnetworkeffectscanhaveamajor impact;Microsoft’sabilitytotapintomultiplenetworkeffectswithWindowsandOfficecontributedgreatly to its unprecedentedly durable franchise. Even today, Windows andOffice remaindominant in thePCmarket; it’s simply thatotherplatforms likemobilehaveachievedsimilarorgreaterimportance.
NetworkEffectsBothProduceandRequireAggressiveGrowth
Akeyelementofleveragingnetworkeffectsistheaggressivepursuitofnetworkgrowth and adoption. Because the impact of network effects increases in asuperlinear fashion, at lower levels of scale, network effects actually exertdownward pressure on user adoption. Once all your friends are on Facebook,youhavetobeonFacebooktoo.Butconversely,whywouldyoujoinFacebookif none of your friends had joined yet? The same is true for the first user of
marketplaceslikeeBayandAirbnb.With network effects businesses, you can’t start small and hope to grow
slowly;untilyourproductiswidelyadoptedinaparticularmarket,itofferslittlevaluetopotentialusers.Economistswouldsaythatthebusinesshastogetpastthe “tipping point” where the demand curve intersects with the supply curve.CompanieslikeUbersubsidizetheircustomersinanattempttomanipulate thedemandcurvetoreachthat tippingpointfaster; thebet is that losingmoneyintheshorttermmayallowyoutomakemoneyinthelongterm,onceyou’repastthetippingpoint.Onechallengethatthisapproachproducesisthe(eventual)needtoeliminate
thesubsidiesinordertomaketheuniteconomicswork.WhenIwasatPayPal,oneofthethingswedidtoencourageadoptionwastoproclaimthattheservicewouldalwaysbefree.Thismeanteatingthetransactioncostsofacceptingcreditcardpayments.IwishIcouldsaywehadagrandplan.Wehadhopedthatwecouldmakeupforthecreditcardtransactionfeesubsidybymakingmoneyoffthe float—the funds being kept in PayPal. Unfortunately, this came nowhereclosetooffsettingthefeesubsidies,andthecompanywashemorrhagingmoney.Sowe switchedPayPal from “always free” to “ACHalways free” and startedchargingfeestoacceptcreditcardpayments.Fortunately,wealreadyhadaloyalfollowing,andourcustomersacceptedthechange.When thebusinesscan’tchange theeconomicsof theproduct (free services
likeFacebookcan’t lower theirprices), itcan insteadswaytheexpectationsofpotentialusers.Thevalueusersplaceontheservicewhendecidingwhetherornot to adopt it depends on both the current level of adoption and theirexpectationsforfutureadoption.Iftheythinkothersaregoingtojumponboard,the perceived value of the service increases, and they becomemore likely toadoptit.Thistechniqueisreflectedinoneofthemostinfluentialbusinessbooksofall
time, Geoffrey Moore’s Crossing the Chasm. Moore argues that technologycompaniesoftenrunintoproblemswhentheytrytotransitionfromamarketofearlyadopterstothemainstream—theproverbial“chasm.”Herecommendsthatcompanies focus on niche beachhead markets, from which the company canexpandoutwardusinga“bowlingpin”strategy inwhich thesemarketshelp toopen up adjacent markets. This strategy is even more important for networkeffectsbusinesses.Acompanycanalsoreshapethedemandcurvebydesigningtheproducttobe
valuabletotheindividualuserregardlessofnetworkadoption.AtLinkedIn,forexample, we discovered that public LinkedIn profiles had some valueindependent of theuser’s network, since they served as anonlineprofessionalidentity. This gave people a reason to join LinkedIn even if their friends andcolleagueshadn’tdonesoyet.
ConnectivityEnablesNetworkEffectsBusinesses
Inadditiontosupportingnetworkeffects,thehighconnectivityoftheworldweliveintodayalsomakesiteasiertoreachthetippingpointwherenetworkeffectskick in, and to sustain those network effects and the market dominance theyproduce.First, the Internethasdriven thecostofdiscovery forproducts and services
lower thanever.Unlike in thepast,whencompaniesneeded tooffergoods inretail storesorbroadcast advertising inorder tobevisible to customers, todaybuyers can find whatever they’re looking for on Amazon or other onlinemarketplaces likeAlibaba, in app stores, or,when all else fails, byGoogling.Becauseproductsandservicesthatarealreadypopularwillalmostalwayscomeup first in search results, companieswith a competitive advantagecanquicklygrow to the point where the increasing returns of network effects produce awinner-take-mostorwinner-take-allmarket.Thisalsoexplainswhythegrowthfactorofdistributionisasormoreimportanttocompanysuccessastheproductitself—withoutdistribution,itisdifficulttoreachthetippingpoint.After network effects take hold, the efficiencies enabled by the Networked
Age make it easier to sustain the pace of rapid growth. In the past, rapidcustomergrowth inevitably led to rapidorganizationalgrowthand todramaticincreases in the overhead required to coordinate a large number of employeesand teams. Today’s networks allow companies to sidestep these traditionalgrowth limiters, suchaswhenAppleusedFoxconn togetaround thepotentiallimitationofitsmanufacturinginfrastructure(moreonthisinthenextsection).Themoreyoucanremovethoselimiters,themoredominantanetworkeffects–driven business can grow. This is why companies like Google that havesurpassed the $100 billion mark in annual revenues are still growing at over20percentperyear.Finally, the remarkable profitability of these companies gives them the
financial resources to expand into new fields and invest in the future. The S-
curve of innovation argues that the rate of adoption of every innovationeventuallyslowsas themarketsaturates.However,companies likeApplehavemastered the strategy of investing in new products that let them hop ontoadditionalS-curves.Applehoppedfrommusicplayerstosmartphonestotablets,anditisnodoubtspendingsomeofitsvastprofitschasingthenextS-curve.Thepremium that the public markets grant these companies also helps them usemergers andacquisitions (M&A) to jump these curves,muchasFacebookdidwithInstagram,WhatsApp,andOculus,andGoogledidwithDeepMind.Ofcourse,networkeffectsdon’tapply toeverycompanyormarket,even if
they are superficially similar—as many companies and their investorsdiscoveredtotheirchagrinduringthedot-combust,theGreatRecession,andthefunding slowdown of 2016. This is why the best entrepreneurs try to designinnovative business models that leverage network effects. One of the reasonsthatGoogleisGoogleandYahoo!isnowpartofAOL(whichinturnisownedby Verizon) is that Google focused on AdWords (a marketplace with strongnetworkeffects)whileYahoo! tried tobecomeamediacompany(a traditionalmodelbasedoneconomiesofscale).MuchofSiliconValley’shistorical success inbuildinggiant companiescan
betracedtoitsculturalemphasisonbusinessmodelinnovation,whichresultsinthecreationofnetworkeffects–drivenbusinesses.TheironyisthatmanypeopleinSiliconValleycouldn’tdefineanetworkeffectorwhatcauseditifasked.Yetsimply because so many entrepreneurs are trying so many different businessmodels, they can end up stumbling into powerful network effects. CraigNewmarksimplystartede-mailinghisfriendsaboutlocaleventsin1995;almosttwenty-twoyearslater,networkeffectshavekeptCraigslistadominantplayerinonlineclassifiedsdespiteoperatingwithaskeletoncrewandmakingseeminglynochangestothewebsitedesignduringthatentireperiod!This is where an emphasis on speed also plays an important role. Because
SiliconValley’sentrepreneursfocusondesigningbusinessmodels thatcangetbig fast, they aremore likely to incorporate network effects.And because thefiercelocalcompetitionforcesstart-upstogrowsoaggressively(i.e.,blitzscale),SiliconValley start-ups aremore likely to reach the tipping point of networkeffectsbeforestart-upsfromlessaggressivegeographies.Oneofthemotivationsforthisbookistohelpentrepreneursfromaroundthe
world emulate these successes by teaching them how to systematically designtheir businesses for blitzscaling. When you design your business model to
leveragenetworkeffects,youcansucceedanywhere.
DESIGNINGTOMAXIMIZEGROWTH:THETWOGROWTHLIMITERS
Buildingkeygrowthfactorsintoyourinnovativebusinessmodelisonlyhalfthebattle.Itisfiendishlydifficulttogrowanamazingbusiness,inpartbecauseitisfiendishly easy to run smack into obstacles that limit your growth. A keycomponent of business model innovation is designing around these growthlimiters.
GROWTHLIMITER#1:LACKOFPRODUCT/MARKETFIT
Product/market fit enables rapid growth, while the lack of it makes growthexpensive and difficult. The concept of product/market fit originates inMarcAndreessen’s seminal blog post “TheOnlyThingThatMatters.” In his essay,Andreessen argues that themost important factor in successful start-ups is thecombinationofmarketandproduct.Hisdefinitioncouldn’tbesimpler:“Product/marketfitmeansbeinginagood
marketwithaproductthatcansatisfythatmarket.”Withoutproduct/marketfit,it’simpossibletogrowastart-upintoasuccessful
business.AsAndreessennotes,
Youseeasurprisingnumberof reallywell-runstart-ups thathaveallaspectsofoperationscompletelybuttoneddown,HRpoliciesinplace, great sales model, thoroughly thought-through marketingplan, great interview processes, outstanding catered food, 30"monitors for all the programmers, top tier VCs on the board—heading straightoff a cliffdue tonot ever findingproduct/marketfit.
Unfortunately,it’sfareasiertodefineproduct/marketfitthanitistoestablishit!Whenyoustartanewcompany,thekeyproduct/marketfitquestionyouneed
to answer is whether you have discovered a nonobvious market opportunitywhereyouhaveauniqueadvantageorapproach,andonethatcompetingplayerswon’tseeuntilyou’vehadachancetobuildahealthylead.It’susuallydifficult
to find such an opportunity in a “hot” space; if an opportunity is obvious toeveryone,thechancethatyou’llbetheonewhosucceedsisexceedinglylow.Most nonobvious opportunities arise from a change in the market that the
incumbents aren’t willing or able to adapt to. In many cases, this can be adisruptive technological innovation, but it can also be a change in the law orfinancial regulations, the riseofanewgroupofcustomers,oranyothermajorshift.Forexample,CharlesSchwabwasable tobuildhiseponymous financialempire by leveraging the deregulation of brokerage commissions to launch adiscountbrokerage.Frequently,youwon’tbeabletofullyvalidateproduct/marketfitbeforeyou
commit to building a company. But you should try. As authors andentrepreneurs,we’rehugefansofEricRiesandhisleanstart-upmethodology.Itisanexcellentprocessforsystematicallytacklingrisk.Butthefactisthatmoststart-upsdon’t followthatprocess; instead, theirchosenexperiment is“Dowesucceedorrunoutofmoney?”Thebestwayforasmall,resource-strappedteamtoassesspotentialstrategies
istoleveragewhatwedubbed“networkintelligence”inourpreviousbook,TheAlliance. Even a small group of founders is likely to have a huge collectivepersonal network of smart people with relevant knowledge or experience.Initiate a conversation, inviting them to challengeyour idea and tell youwhatelseyoushouldconsider.Of course, even the best network intelligence won’t guarantee that you’ve
actuallyfoundproduct/marketfitduringthisdesignphase.Theonlywaytotrulyproveproduct/market fit is toget theproduct into thehandsof realusers.Butentrepreneurscanandshoulddotheirresearch,andtrytodesigntheirbusinessmodel tomaximize theirchancesofachievingproduct/market fitasquicklyaspossible.
GROWTHLIMITER#2:OPERATIONALSCALABILITY
Designing a scalable economicmodel isn’t enough if you can’t scale up youroperations tomeetdemand.Toooften,entrepreneursdismiss thechallengesofoperational scalability by saying, “Managing explosive growth is a high-classproblem.”High-classproblemsarestillproblems;itmayfeelbetterforyouregoto be wrestling with the issues of growth rather than simply trying to avoidmissingpayroll,butbothcanstillkillyourcompany.Ratherthandismissthese
challenges,thewisestinnovatorsdesignoperationalscalabilityintotheirmodels.
A)HumanLimitationsonOperationalScalability
A significant number of operational issues arise simply because of humanlimitations.Asmuchaswemightwish thatweandourcolleaguescouldworktirelesslyandseamlessly,regardlessof thescaleof theorganization, thefact isthatgrowthcausesustotripoverawidearrayofissues.If you are leading a small founding team with four members, you have to
worryaboutyourdirect relationshipwith the threeothercofounders,plus theirdirect relationships with one another. Combinatorial mathematics tells us thatthismeansyouneedtomanagetherelationshipsbetweensixpairsofindividuals([4*3] /2).Nowimagine thatyouhire twoemployees, fora total teamsizeofsix.Nowyouneedtomanagetherelationshipsbetweenfifteenpairs([6*5]/2).Youincreasedtheteamsizeby50percent,butthenumberofrelationshipsyouneedtomanagewentupby150percent.Themathjustgetsmoredauntingfromthere. And that only considers the relationships of individual pairs of teammembers,nottherelationshipsbetweenanythreemembers,anyfourmembers,andsoon.Oneapproachistodesignabusinessmodelthatrequiresasfewhumanbeings
aspossible.Somesoftwarecompaniesemploybusinessmodelsthatallowthemtoachievemassivesuccesswithminimalnumbersofemployees.ThefoundersofWhatsApp,JanKoumandBrianActon,designedacleverbusinessmodelthataddressed some of the key growth factors (their messaging service leveragedbothclassicnetworkeffects and the existingdistributionnetworkof telephoneaddress books to grow faster) but also managed to skirt around issues ofoperational scalability.WhatsApphad a freemiumbusinessmodel; the servicewas free for a year, after which it cost $1 per year. This low-friction modelessentially eliminated the need for people working in functions like sales,marketing, and customer service, allowingWhatsApp to grow to five hundredmillionmonthlyactiveusersbythetimeofitsacquisitionbyFacebook,withastaff of just forty-three employees, a ratio of over tenmillion active users peremployee!Another approach is to find ways to outsource work to contractors or
suppliers. Airbnb’s strategy for photographing its hosts’ rooms offers aninstructiveexample.Earlyon,Airbnb’sfoundersdiscoveredthatoneofthekey
factorsthatincreasedthechancesofrentingaroomonAirbnbwasthequalityofthe photographs of that room. It turns out that most of us aren’t professionalphotographers,andourpoorlycomposed,poorlyshotcellphonepicturesdon’tdo a good job of conveying the awesomeness of our living spaces. So thefounders took to the road, visiting hosts and taking photographs for them.Obviously,personallyvisitingeveryhostwashardlyascalablesolution,sothethe taskwas soonoutsourced to freelancephotographers.AsAirbnbgrew, thestrategyshiftedfromthefoundersmanagingashortlistofphotographers,toanemployee managing a large group of photographers, to an automated systemmanagingaglobalnetworkofphotographers.FounderBrianCheskydescribesthis strategy succinctly: “Do everything by hand until it’s too painful, thenautomateit.”Ultimately, evenwith clever businessmodels and automation, nearly every
massivelysuccessfulcompanyrequires thousandsoreven tensof thousandsofemployees.Smarttechniquescandelaythereckoning,butnotforever.Lateron,we’ll discuss some of the management innovations that allow companies tohandlethiskindoforganizationalgrowthandscale.
B)InfrastructureLimitationsonOperationalScalability
The other main challenge of operational scalability comes from the strain ofscaling up the nonhuman infrastructure of the business. It doesn’tmatter howmuchdemandyougenerate ifyour infrastructurecan’thandle it. Infrastructurelimitationscanevenbefatal toacompany’sambitions.Consider theexamplesofthesocialnetworksFriendsterandTwitter.While many have forgotten it now, Friendster was the first (pre-Facebook)
onlinesocialnetworktobreakthroughintothemainstream(disclosure:Iwasanearly investor in Friendster). Launched in March 2003, Friendster rode viralgrowthtomillionsofuserswithinmonths.Beforetheyearwasout,Friendster-maniawassuchaculturalphenomenonthatfounderJonathanAbramsappearedon the late-night television program Jimmy Kimmel Live! But Friendster’smassivegrowthbroughtmassiveheadaches,especiallyontheinfrastructureside.Despite a talented technology team, Friendster’s servers couldn’t handle thegrowth,anditbecamecommonforFriendsterprofilestotakeuptofortysecondsto load. By the beginning of 2005, a faster new entrant, MySpace, wasgeneratingmore than ten times the number of pageviews as Friendster,which
never recovered. MySpace, of course, ultimately lost the consumer socialnetworkingwartoFacebook,whichisastorywe’lldiscussindetaillaterinthisbook.Twittercameclosetomeltingdowninthesameway,butmanagedtorecover
in time to build a massive business. When Twitter began its rise in the late2000s,itbecameinfamousforits“FailWhale,”awhimsicalerrormessagethatappeared whenever its servers couldn’t handle the load. Unfortunately forTwitter, the FailWhale made fairly regular appearances, especially when bignewshit,suchasthedeathoftherecordingartistMichaelJacksonin2009(tobefair,TwitterwashardlytheonlywebsitethathadtheseissueswhentheKingofPop passed away) or the 2010World Cup. Twitter invested serious resourcesinto rearchitecting both its systems and its engineering processes to be moreefficient.Evenwiththisstrenuouseffort,ittookseveralyearsto“tame”theFailWhale; it wasn’t until after Twittermade it through the 2012US presidentialelectionnightwithoutmeltingdown that the company’s then–creativedirectorDougBowmanannouncedthattheGreatBlueWhalehadbeenputtodeath.Oneofthemainreasonsfortheverylargeincreaseinthegrowthofvaluable
Web companies that we’ve seen in recent years is Amazon’s cloud offering,AmazonWebServices(AWS),whichhashelpedmanysuchbusinessesnavigatearoundinfrastructurelimitations.Dropbox,forexample,wasabletoscaleupitsstorage infrastructure much more quickly and easily because it used AWSstorage,eliminatingtheneedtobuildandmaintainitsownarraysofharddisks.AWSreflectsoneofthewaysthatAmazonhasmadeoperationalscalabilitya
competitiveadvantage.WebserviceslikeAWStapintowhatHarvardBusinessSchoolprofessorCarlissBaldwinandformerHarvardBusinessSchoolprofessorKimClarkrefertoas“thepowerofmodularity.”AsBaldwinandClarkdescribein their book,Design Rules, Vol 1: The Power of Modularity, this principlemakes it possible for a company like Amazon and its customers to buildcomplex products out of smaller, standardized subsystems. But the power ofmodularitygoesbeyondjustsoftwaredevelopmentandengineering.Bybuildingeasy-to-integrate subsystems like payments and logistics, Amazon makes itsentirebusinessmoreflexibleandrapidlyadaptable.TheequivalenttoAWSonthehardwaresideisChina.Hardwarestart-upsare
able to manage infrastructure limitations and scale much more quickly bytapping into Chinesemanufacturing capabilities, either directly or byworkingwith companies like the custom manufacturing design firm PCH. The smart
thermostat maker Nest, for example, had only 130 employees when it wasacquired byGoogle for $3 billion, largely because it had outsourced all of itsmanufacturingtoChina.In contrast, Tesla Motors has seen its growth held back by infrastructure
limitations. Due to the complexities of its manufacturing process, Tesla’sproductionrateshavelaggedbehindthoseofotherautomakers,theresultbeingthat its award-winning vehicles are almost always sold out, with back ordersmeasured inmonths and even years.Demand generation is not a problem forTesla;meetingthatdemandis.
PROVENBUSINESSMODELPATTERNS
Whether by design or not, the businessmodels of rapidly growing companiesoften follow proven patterns that tap into growth factors and bypass growthlimiters.Thesepatternswillbedescribedinmoredetailbelow,buthereitbearsnoting that these high-level patterns are principles rather than exact recipes.Simply adopting any of these particular patterns isn’t enough to ensure aninnovativebusinessmodel,butunderstandingthemdoesprovideanentrepreneurwithasetofgoodrolemodels.It is also worth mentioning that not all patterns are created equal. Some
commonbusinessmodelsfollowprovenpatterns,butnonethelessdon’tseemtoproduce $100 billion businesses or even $10 billion businesses. Take open-source software, which has been wildly successful as a pattern for spreadingsoftware products like Linux. Open source, which means offering free,community-createdsoftwarethatuserscanmodify,arosetoprominenceduringthe dot-comera andhas been an integral part of theworld’s technology stackeversince.The story of open-source software fits the pattern of business model
innovation. Open-source software serves a large market, has powerfuldistribution via open-source software code repositories, benefits from thenetwork effects of standards and compatibility, andneatly avoidsmanyof thehuman limitations on operational scalability by tapping into a distributedcommunityofvolunteercontributorsratherthanbuildingalargeorganizationofemployees.Yet even themost successful open-source business, RedHat, has amarket
capitalizationof“only”about$15billion,andthat’safterbeinginbusinessfor
twodecades.Theempiricalevidencesuggeststhatopensourceisapatternthatisvaluableforengagementbutnotforbuildingamassivelyprofitablebusiness.Inorderforapatterntobeproven,itmustbeabletodemonstratethatmultiple
massively valuable businesses follow it. Based on that criterion, we’veassembledthefollowinglistofprovenpatternstohelpinspireyourownbusinessmodelinnovation.
PROVENPATTERN#1:BITSRATHERTHANATOMS
Google and Facebook are largely software businesses that focus on electronicbits rather thanmaterial atoms.Bits-basedbusinesseshaveamucheasier timeservingaglobalmarket,whichinturnmakesiteasiertoachievealargemarketsize.Bitsarealsofareasiertomovearoundthanatoms,sobits-basedbusinessescanmoreeasilytapintodistributiontechniqueslikevirality,andtheirabilitytobe highly networked providesmore opportunities to leverage network effects.Bits-based businesses tend to be high-gross-margin businesses because theyhavefewervariablecosts.Bitsalsomakeiteasiertodesignaroundgrowthlimiters.Youcaniteratemore
quickly on software products (many Internet companies release new softwaredaily) than on physical products, making it faster and cheaper to achieveproduct/market fit.Andbits-basedbusinesses, aswe sawwithWhatsApp, cangetawaywithfarfeweremployeesthanmostoftheiratom-basedcounterparts.Backin1990,thefuturistGeorgeGilderdemonstratedhispresciencewhenhe
wroteinhisbookMicrocosm,“Thecentraleventofthetwentiethcenturyistheoverthrow of matter. In technology, economics, and the politics of nations,wealth in the form of physical resources is steadily declining in value andsignificance.Thepowersofmindareeverywhereascendantoverthebruteforceofthings.”Just over twenty years later, in 2011, the venture capitalist (and Netscape
cofounder)MarcAndreessenvalidatedGilder’sthesisinhisWallStreetJournalop-ed “Why Software Is Eating theWorld.” Andreessen pointed out that theworld’s largest bookstore (Amazon), video provider (Netflix), recruiter(LinkedIn), and music companies (Apple/Spotify/Pandora) were softwarecompanies,andthateven“oldeconomy”stalwartslikeWalmartandFedExusedsoftware(ratherthan“things”)todrivetheirbusinesses.
Despite—orperhapsbecauseof—thegrowingdominanceofbits, thepowerof softwarehas alsomade it easier to scaleup atom-basedbusinesses aswell.Amazon’s retail business is heavily based in atoms—just think of all thoseAmazon shipping boxes piled up in your recycling bin! Amazon originallyoutsourced its logistics to IngramBookCompany, but its heavy investment ininventorymanagementsystemsandwarehousesasitgrewturnedinfrastructurelimitations from a growth limiter to a growth factor. On the retail side,merchantspayAmazontomanagetheirinventoriesandlogisticsforthem,whilethemassive computer systems thatAmazon built to operate its retail businessgaveitthecapabilitiestolaunchitsAWSbusiness(whichisahigh-margin,bits-basedbusiness!).
PROVENPATTERN#2:PLATFORMS
Platformeconomicspredates theNetworkedAge,andeven the IndustrialAge.Trade-orientedprincipalitiesliketheRepublicofVeniceprovidedawelcomingecosystemformerchants,completewithcurrencyandtheruleoflaw,aswellastaxestoharvestthevalueoftheplatform.TechnologyplatformslikeMicrosoftWindows demonstrated the power of being the chosen platform on whichbusinesseswere built backwhen theWorldWideWebwas still a glimmer inTim Berners-Lee’s eye (Sir Berners-Lee wrote his proposal for a globalhypertextsystemin1989).Yetdespitetheprovenvalueofplatformsinthepre-Internet era, the Networked Age has made them vastly more powerful andvaluable.RatherthanbeinglimitedliketheRepublicofVenicetoaspecificgeography,
today’s software-based platforms can achieve global distribution almostimmediately.Andsincetransactionsontoday’splatformsareconductedthroughapplication programming interfaces (APIs) rather than person-to-personnegotiations, they proceed swiftly, seamlessly, and in incredible volumes, allwithbarelyanyhumanintervention.Ifaplatformachievesscaleandbecomesthedefactostandardforitsindustry,
thenetworkeffectsofcompatibilityandstandards(combinedwiththeabilitytorapidly iterate and optimize the platform) create a significant and lastingcompetitiveadvantage thatcanbenearlyunassailable.Thisdominance lets themarket leader“tax”all theparticipantswhowant touse theplatform,muchaslevieswereimposedinthebygoneRepublicofVenice.Forexample,theiTunes
storetakesa30percentshareoftheproceedswheneverasong,amovie,abook,or an app is soldon that platform.Theseplatform revenues tend tohaveveryhighgrossmargins,whichgeneratecash thatcanbeplowedback intomakingtheplatformevenbetter.Amazon’smerchantplatform,Facebook’ssocialgraph,and, of course, Apple’s iOS ecosystem are great examples of the power ofplatforms.
PROVENPATTERN#3:FREEORFREEMIUM
“Free”hasanincrediblepowerthatnootherpricingdoes.TheDukebehavioraleconomist Dan Ariely wrote about the power of free in his excellent bookPredictably Irrational, describing an experiment inwhich he offered researchsubjectsthechoiceofaLindtchocolatetrufflefor15centsoraHershey’sKissforamerepenny.Nearlythree-fourthsofthesubjectschosethepremiumtrufflerather than the humbleKiss.ButwhenAriely changed the pricing so that thetruffle cost14cents and theKisswas free—the samepricedifferential—morethantwo-thirdsofthesubjectschosetheinferior(butfree)Kisses.The incredible power of free makes it a valuable tool for distribution and
virality. It also plays an important role in jump-starting network effects byhelping a product achieve the criticalmass of users that is required for thoseeffectstokickin.AtLinkedIn,weknewthatourbasicaccountshadtobefreeifwewantedtogettothemillionuserswetheorizedrepresentedcriticalmass.Sometimes you can offer a product for free and still be profitable; in the
advertising-driven business model, a large enough mass of free users can bevaluableeveniftheyneverpayforyourservice.Facebook,forexample,doesn’tcharge itsusersadime,but it is able togenerate largeamountsofhigh-gross-marginrevenuebysellingtargetedadvertising.Butsometimesaproductdoesn’tlend itself to theadvertisingmodel, as is thecasewithmany servicesusedbystudentsandeducators.Without third-party revenue, theproblemwithofferingyour product to users for free is that you can’t offset your lack of sales by“makingitupinvolume.”Here is where the innovation of freemium comes in. The venture capitalist
FredWilsoncoined the term ina2006blogpost (basedona suggestion fromJaridLukin),butthebusinessmodelitselfpredatestheterm,havingitsorigininthe“shareware”modelforsellingsoftwareinthe1980s.Thefreeproductwasatoolfordiscoveryandgainingacriticalmassofusers,whilethepaidversionof
thesoftwareallowsthebusinesstoextractvaluefromthoseusersonceitsvalueis clear. Dropbox is one of the premier examples of a successful freemiumbusiness—by giving away 2GBof storage,Dropbox attracted amassive userbase, a reasonable percentage of which decides to pay for the value andconvenienceofadditionalstorage.
PROVENPATTERN#4:MARKETPLACES
Marketplacesrepresentoneofthemostsuccessfulbusinessmodelpatterns,withthedot-comera’sGoogleandeBayandtoday’sAlibabaandAirbnbstandingoutas examples of important, valuable companies that follow this pattern. Onereason marketplaces are powerful is because they often tap into two-sidednetwork effects.While it is difficult to create a successfulmarketplace fromacold start, the first marketplace that does manage to achieve liquidity—theability for buyers and sellers to quickly and efficiently find a counterparty toconducta transaction—becomesveryattractive tobothsidesof themarket.Asbuyersandsellerspourin,themarketplacebecomesevenmoreattractivetobothparties, triggering a positive feedback loop that makes it very hard for newentrantstowinanymarketshare.Marketplacesalsoofferkeyadvantagesbeyond theobviousnetworkeffects.
By creating a liquid market where buyers and sellers both participate, thedynamic forces of supply and demand price their transactions better than anyhumanjudgmentcould.Themoreefficientthepricesinamarketplace,themorevalue it creates, because thatmeansmore transactions thatmight create valueactually occur. In contrast, in illiquid markets, sellers often misprice theirproducts,resultinginfewersalesandlessvaluecreationthanoptimal.The best example of the benefits of efficient market pricing is probably
Google’sAdWordsadvertisingmarketplace.AdWordsallowsanyonetobidontargetedkeywords,inanyquantity,soeventhesmallestbusinessescantapintoglobaldistribution.Contrast this to the traditionaladvertisingmarket, inwhichlarge clients spend millions of dollars paying advertising agencies to runexpensive thirty-second television ads during coveted programming like theSuperBowlbroadcast.Google’s systemalsomeasures advertisingquality; adstargetedatitsaudiencetogeneratethemostpaidclick-throughsarefavored.Theneteffectisthatconsumersareshownthemosteffectivelytargetedads,withoutthe overhead of a middleman like Don Draper and his three-martini lunch.
Googlealsoincreasesitsowngrossmargin,because,unlikecommercialsduringa television broadcast, search-based ad space is virtually unlimited and costsGooglenexttonothing.Although marketplaces, even local ones, have always been a powerful
businessmodel,thechangesusheredinbytheNetworkedAgehavemadethempotentially more valuable than ever. But unlike a local market with its sizeconstraints—thinkofanold-fashionedbazaarinthecenterofapopulouscity—onlinemarketplaces tapaglobalmarket.Andbyconnectingbuyersandsellersinstead of holding inventory or managing logistics (and thus dealing in bitsrather than atoms), online marketplaces avoid many of the growth limits ofhumanorinfrastructurescalability.
PROVENPATTERN#5:SUBSCRIPTIONS
When Salesforce.com first launched its on-demand customer relationshipmanagement product, there were many legitimate questions about this newsoftware-as-a-service(SaaS)model.Sellingsoftwareasasubscription,deliveredvia the Internet, representedamajordeparture forenterprise softwarevendors.Thepreviousmodelof sellingpermanent licenses foron-premisesoftwareandcharging formaintenanceprovidedmorecashup front thanmonthlyorannualsubscriptions.Thepersonnelrequiredtosupport themodelwerealsodifferent;sellingandsupportingon-premisesoftwarerequiredfieldsalespeopleandsalesengineers to install pilot deployments, while the new SaaS model requiredadditionalstafftoprovide24/7datacentercoverageandsupport.As it turns out, of course, SaaS eventually became the dominant business
model for enterprise software. The cash flow disadvantages and requiredpersonnel shifts were real concerns, but mainly for existing players in themarket.NewSaaSbusinesses likeSalesforce.comandWorkdayweredesignedandbuilt around thenewmodel, giving themamajor advantageover existingplayerswhotriedtoconverttheiron-premisesoftwarebusinessestosubscriptionones.Subscription Internet services have been successful because the sales and
delivery model provides a larger market size and better distribution thantraditional packaged software. Due to the cost and overhead of the extensivefield operations required to support on-premise software, traditional enterprisesoftwarelicenseshadtobeinthesix-orseven-figurerangesimplytomakethe
modelwork.Thismeantthatsoftwarevendorsfocusedontheneedsofonlythelargestcustomers.Incontrast,Salesforce.comandotherSaaSvendorscansellsoftwarelicenses
inanyquantity,notonlytoFortune500companies,butalsotomidmarketandsmalltomedium-sizedbusinesses,significantlyenlargingtheirpotentialmarket.Internet delivery and self-service allownew forms of distribution thatweren’tpossible in the packaged softwareworld, such asDropbox’s viral incentive ofadditionalfreestorageforreferringnewcustomers.NoristhepatternofInternetsubscriptionslimitedtoenterprisesoftware.The
dominant players in both music (Spotify, Pandora) and video (Netflix, Hulu,Amazon) also enjoy lower overhead and greater distribution by using thesubscriptionbusinessmodel.Another,lessobviousbenefittothismodelisthatonceasubscriptionbusiness
achieves scale, the predictability of its revenue streams allows it to be moreaggressivewith long-term investments, since it isn’t obliged tomaintain largecash balances to weather short-term variations in the business. This financialfirepower can represent amajor competitive advantage. For example, Netflix,whichannouncedplanstoinvest$6billioninoriginalcontentforitsstreamingservice in2017,hasexploited itsdirect subscriptionmodel tooutspendclassictelevision networks, which have to rely on less robust revenue streams likepaymentsfromcableprovidersandadvertisingsales.
PROVENPATTERN#6:DIGITALGOODS
One of the emerging patterns that build on new platforms and services is thebusinessof sellingdigitalgoods.Sittingat the intersectionof “bits rather thanatoms”andplatforms,digitalgoodsareintangibleproductsthat,arguably,havenointrinsicvalue—buttheycanstillmakeforaprofitableandscalablebusiness.Forexample,themessagingserviceLINEderivessignificantrevenuesbyselling“stickers”:imagesthatareincorporatedintothetextofsmartphonemessages.In2014,itsfirstyearofoperation,LINE’sstickerbusinessgenerated$75millioninrevenue. That figure grew to $270million in 2015, which represented over aquarter of LINE’s total revenues. Not bad for an intangible product with nointrinsicvalue!Digital goods have also become a key business model in the video game
industry,within-gamepurchasesofdigitalitemsthatcanhelpplayersadvance
inthegameoradvertisetheirstatus.Market-widerevenuefromin-apppurchasesareprojectedtooutstrippaid-appdownloadsin2017,$37billionto$29billion.In addition to enjoying the advantages of any bits-based business, digital
goods tend to have nearly 100 percent gross margins, since they are purelydigitalandusuallydonotaddsignificantlytoinfrastructureoroverheadcosts.
PROVENPATTERN#7:FEEDS
One of themost underrated and underappreciated proven patterns is the newsfeed.Facebook’spowerfulnetworkeffectsallowthesitetoattractitsusers,butitsinnovationofthenewsfeedhasmadeitaworld-classbusiness.YetFacebookishardlytheonlyfeed-centricsuccessstory.CompanieslikeTwitter,Instagram,andSlackhaveallbuiltmultibillion-dollarmarketvaluesaroundthenewsfeedpattern.Thepowerofthenewsfeedcomesfromitsabilitytodriveuserengagement,
which in turn drives both advertising revenue and long-term retention. AsFacebook has demonstrated, a news feed with sponsored updates is the mosteffective way to monetize proverbial Internet “eyeballs.” Facebook’s NewsFeed’sdominanceoftheonlineadvertisingmarketisonlyexceededbyGoogle’sAdWords, and AdWords starts with the significant built-in advantage ofcapturingactiveconsumerintentratherthansimplythedesiretobeamused.Forexample,howmanypeoplevisitFacebookwiththeintentionofgoingshopping?Themagicofthenewsfeedmodelhasbeenitsabilitytomonetizeboredpeoplecatchinguponwhattheirfriendsaredoing.Ofcourse,effectiveuseofthenewsfeedmodelrequiresalotofsophisticated
technology. Facebook doesn’t just insert sponsored updates at random. Thecompanyknowsyourinterestsbetterthanyoudo,basedonalltheitemsyou’veever clicked on, liked, or otherwise engaged with. It can carefully target theadvertisementsitshowsyoubasedonyourindividualhabitsandthecontextofwhatsurroundstheminyourfeed.ThistargetingabilityexplainswhyFacebooksucceeded inmonetizing thismodelwhen other feed-based products likeRSSreadersfailed.ThispatternissopowerfulthatTwitter,whoseproductisessentiallyonelong
news feed, is still an important Internet company despite barely changing itsproductinnearlyadecade(goingfrom140charactersto280charactersdoesn’tcount). Twitter is a business that scaled massively because of the power of
businessmodelinnovation,notproductortechnologyinnovation.
THEUNDERLYINGPRINCIPLESOFBUSINESSMODELINNOVATION
Underlying the proven patterns of business model innovation are largerprinciples that can help refine those patterns or even create new ones. Theseprinciples aren’t themselves business models, but they often power thetechnologicalinnovationthatenablesbusinessmodelinnovation.
UNDERLYINGPRINCIPLE#1:MOORE’SLAW
Moore’s Law is the fundamental principle that puts the “Silicon” in SiliconValley, and has powered the worldwide ascent of the technology industry.Moore’s Law is named after its codifier, Intel cofounderGordonMoore,whocoined the term in a paper he wrote in 1965, observing that the number oftransistorsthatcouldbecrammedontothesurfaceofasiliconchipappearedtodouble each year. While Moore revised his eponymous law in 1975 to adoublingoftransistorseverytwenty-fourmonths, theindustryhassincesettledonabroadconsensusofeighteenmonths.Today,Moore’sLawnolongerrefersspecificallytotransistordensity;rather,itpredictsthatcomputingpowertendstodoubleeveryeighteenmonths.Inrecentyears,thisgrowthincomputingpowerhasbeendrivenbythetransitiontomulticore,multithreadedcomputing.Perhapsinthefuture,Moore’sLawwillbemetbyquantumcomputing,opticalchips,theuse of DNA, or something even more impossible to foresee. The point is, itappearsthatthetruelimittoMoore’sLawishumanengineeringingenuity,notsolid-statephysics.Moore’sLawmattersbecausetherelentlessincreaseincomputingpowerthat
it predicts acts as a constant source of technological innovation,which, aswehave seen, can help enable business model innovation. For many years, thepowerof Intel’scentralprocessingunits (CPUs)wasmeasuredby their“clockrate”—thenumberoftimespersecondthattheCPUcouldperformanoperation.Whileclock rate isno longeragoodmeasureofcomputingpower, it is still agoodmetaphorforhowMoore’sLawdrivestheworldofcomputertechnology:each tick of the clock enables new technologies, driving faster and fasterinnovations.Increasing computing power allowed the shift from gigantic mainframes to
smaller minicomputers to personal computers, all the way to today’ssmartphonesandwearables.We’veseensimilarincreasesinthingslikenetworkbandwidth,allowingtheWebtoshiftfromtexttoimagestoaudiotovideo,andin the future, 3-D and virtual reality (VR). Yet today’s smartphones aren’tsimplysmallerversionsofIBMmainframes—remember,technologyinnovationenablesbusinessmodelinnovation.The best entrepreneurs don’t just follow Moore’s Law; they anticipate it.
Consider ReedHastings, the cofounder and CEO ofNetflix.When he startedNetflix,hislong-termvisionwastoprovidetelevisionondemand,deliveredviatheInternet.Butbackin1997,thetechnologysimplywasn’treadyforhisvision—remember, this was during the era of dial-up Internet access. One hour ofhigh-definition video requires transmitting 40 GB of compressed data (over400GBwithout compression).A standard 28.8Kmodem from that erawouldhave taken over fourmonths to transmit a single episode of Stranger Things.However, therewasa technological innovation thatwouldallowNetflix togetpartwaytoHastings’sultimatevision—theDVD.Hastings realized thatmovieDVDs, then selling for around $20,were both
compact and durable. This made them perfect for running a movie-rental-by-mail business.Hastings has said that he got the idea froma computer scienceclassinwhichoneoftheassignmentswastocalculatethebandwidthofastationwagonfullofbackuptapesdrivingacrossthecountry!Thiswastrulyacaseoftechnological innovation enabling business model innovation. BlockbusterVideohadbuiltasuccessfulbusinessaroundbuyingVHStapesforaround$100andrentingthemoutfromphysicalstores,butthebulky,expensive,fragiletapeswouldneverhavesupportedarental-by-mailbusiness.(As hard as it may be for some readers to comprehend, when we were in
college, we would often drive to a Blockbuster Video store on a Friday orSaturdaynight,payacoupleofbuckstorentaVHStapeofamovie,andusealandline telephone to call Domino’s to order a pizza before popping thevideotape into a VCR that was connected to a twenty-five-inch standard-definitioncathode-raytube.)DVDtechnologyallowedNetflixtocreateacompletelynewbusinessmodel.
Ratherthanrentingoutindividualmoviesandbeingchargedexorbitantlatefeesif they failed to return theVHS tape in time,Netflix customers paid $20 permonthforasubscriptionto“unlimited”movies—providedtheycheckedoutjustone movie at a time. This allowed Netflix to eliminate Blockbuster’s widely
loathedlatefeesandcapture thepowerfulandcertainrevenuestreamfromtheprovenmodelofasubscriptionservice.Netflixtookoff,andevenwentpublicasaDVD-by-mailservice.ButHastingsnever lost sight of his ultimatevision forNetflix—on-demand
televisiondeliveredviatheInternet.AndasMoore’sLawcontinuedtoworkitsmagic, making computers ever more powerful and Internet bandwidth evergreater and cheaper, Netflix bided its time, waiting for streaming video tobecomeviable.“Whenwe first started raisingmoney in 1997, we thought we’d bemostly
streamingin5years,”HastingstolduswhenhevisitedourBlitzscalingclassatStanford.“In2002,wehadnostreaming.Sowethoughtthatby2007,itwouldbe half our business. In 2007, we were still nowhere. So we made the sameprediction. And this timewewerewrong the otherway—by 2012, streamingwas60%ofourbusiness.”ItmayhavetakenlongerthanHastingsexpected,butMoore’sLaweventuallycamethroughforhim.Today, Netflix is synonymous with television on demand delivered via the
Internet,andithascreatedanentirelynewcategoryof“bingewatching.”Asof2017, 53 percent of American adults say that their household has access toNetflix,andtheservice isgrowingrapidlyacross therestof theworld.Netflixhas used the financial power of its subscription model to become one of thepremier sourcesoforiginal video content, from television shows likeStrangerThings, to movies like Beasts of No Nation, to events like comedian DaveChappelle’scomebackstand-upcomedyspecials.Traditional television commissions large numbers of pilot episodes, the
majorityofwhichnevermakeittoseries,tryingtoproduceoptimisticallynamed“MustSeeTV”toappealtoabroadaudience,whichhastobeconvincedtotuneineverysingleweek.Incontrast,theon-demandmodelallowsNetflixtocatertomany different audiences rather than program a small number of thematicchannels,ascabletelevisiondoes.Broadcasttelevisionsucceededbyprovidingthe same thing to all its viewers—a model driven by the technologicalinnovationofbroadcastingcontentviawireless signalsand latercoaxialcable.Netflixsucceedsbyprovidingacarefullypersonalizedexperiencetoeachofitsmany viewers, giving it a huge advantage over its traditional televisioncompetitors. Moreover, Netflix produces exactly what it knows its customerswantbasedontheirpastviewinghabits,eliminatingthewasteofallthosepilots,andonly loses customerswhen theymake a proactive decision to cancel their
subscription.ThemoreapersonusesNetflix,thebetterNetflixgetsatprovidingexactly what that person wants. And increasingly, what people want is theoriginalcontentthatisexclusivetoNetflix.ThelegendaryscreenwriterWilliamGoldmanfamouslywroteofHollywood,“Nobodyknowsanything.”TowhichReedHastingsreplies,“Netflixdoes.”AndallthiscameaboutbecauseHastingshadtheinsightandpersistencetowaitnearlyadecadeforMoore’sLawtoturnhis long-term vision from an impossible pipe dream into one of the mostsuccessfulmediacompaniesinhistory.Moore’s Law has worked its magic many other times, enabling new
technologies ranging from computer animation (Pixar) to online file storage(Dropbox) to smartphones (Apple). Each of those technologies followed thesamepath frompipe dream toworld-conquering reality, all driven byGordonMoore’s1965insight.
UNDERLYINGPRINCIPLE#2:AUTOMATION
Blitzscalingcompaniesuseautomation.Iftheyhavetheabilitytoperformatask(which is a big if), computers are almost always faster, cheaper, and morereliable thanhumanbeings.Furthermore, computers continue toget faster andcheaper,doublinginpowereveryeighteenmonthsaccordingtoMoore’sLaw,asopposed to human beings, who evolve over the course of millions of yearsaccordingtoDarwin’sprincipleofnaturalselection.In 2014, the journalist Jan Vermeulen compared the original Apple II
(introduced in1977)with the thenstate-of-the-art iPhone5S.He found that inthe intervening thirty-seven years, Apple’s products had become 2,600 timesfaster in terms of clock speed (from a 1MHz single-coreCPU to a 1.3GHzdual-coreCPU)andhad16,384timestheamountofRAM.That’sthreetofourordersofmagnitudeof improvement in thespanofasinglehumangeneration.And thatmassivedeltadoesn’t even take intoaccount that theApple IIwasadesktopcomputerwithabulkycathode-raytubemonitor,andtheiPhone5Swasaportablesupercomputerthatpeoplecarriedintheirpockets.The same year that the Apple II was introduced, Joe Bottom set a world
recordbyswimmingthe50-meterfreestylein23.74seconds,forabriskpaceofjust under 7.6 km/h (4.7 mph). If human swimming speed had increased asquicklyas thecomputing speed inApple’sproducts, theworld record in2014would have been 19,700 km/h (12,250 mph)—not quite enough to achieve
orbitalvelocity,butabouttwenty-fivetimesthespeedoftheaveragecommercialjetliner.Theactualhumanworldrecordfor the50-meterfreestyle in2014was20.91seconds,foramoremodest11percentimprovement.That’sthepowerthatautomationtapsinto.Thepowerofautomationappliesnotjusttodirect-to-consumerproductslike
theiPhonebutalsotointernalprocessesandcapabilities.Thinkofthevaluethatautomation creates by increasing the productivity inAmazon’swarehouses, orbymakingiteasiertokeepGoogle’sserverfarmsrunning24/7.
UNDERLYINGPRINCIPLE#3:ADAPTATION,NOTOPTIMIZATION
At a higher level of abstraction, successful scale-ups placemore emphasis onadaptation than optimization. Rather than the giant assembly lines of Detroitautomakers, which trace their origins to Henry Ford’s Model T, the currentgeneration of Silicon Valley companies practice continuous improvement,whether through an emphasis on speed or the constant experiments and A/Btestingofgrowthhacking.Thisemphasismakessenseinanenvironmentwherecompanies need to seek product/market fit for new and rapidly changingproducts andmarkets.Consider howAmazon expanded into newmarkets likeAWSratherthansimplyhoningitsretailcapabilities,orhowFacebookhasbeenabletoadapttotheshiftfromatext-basedsocialnetworkaccessedviadesktopWeb browsers to an image-and video-based social network accessed viasmartphones(andsoon,perhaps,VR).
UNDERLYINGPRINCIPLE#4:THECONTRARIANPRINCIPLE
My friend Peter Thiel has written eloquently about the power of being acontrarianinhisbookZerotoOne.
WheneverIinterviewsomeoneforajob,Iliketoaskthisquestion:“Whatimportanttruthdoveryfewpeopleagreewithyouon?”Thisquestionsoundseasybecauseit’sstraightforward.Actually,
it’s very hard to answer. It’s intellectually difficult because theknowledgethateveryoneistaughtinschoolisbydefinitionagreedupon. And it’s psychologically difficult because anyone trying to
answermust say something she knows to be unpopular. Brilliantthinkingisrare,butcourageisinevenshortersupplythangenius.
Being contrarian is often critical to the process of creating a massivelyvaluable technology company. As we’ve discussed, key growth factors likedistribution and network effects tend to provide disproportionate rewards to acompany that is the first in its space toachievecritical scale.Beingcontrarianandrightgivesyouahugeadvantagebecauseyougetaheadstartonachievingscale.If your company is pursuing an opportunity that nearly everyone agrees is
very attractive, you’re likely to have a difficult time distancing yourself fromyourarmyofcompetitors.Butifyourcompanyispursuinganopportunitythatconventionalwisdomignoresordisdains,youwillprobablyhave the timeyouneed to refine your business model innovation into a well-oiled machine.Amazonpursuede-commercewhenmostpeopledidn’tthinkconsumerswouldfeel comfortable using credit cards online. Google launched its search enginewhenmostpeoplethoughtsearchwasamaturecommodity.AndFacebookbuiltits social network when most people believed social networking to be eitheruseless,amarketdominatedbyMySpace,orboth.Aswe’vealreadyseen,mostgreatideaslookdumbatfirst.Beingcontrarian
doesn’tmean that dumbpeople disagreewith you; itmeans that smart peopledisagreewithyou!RememberwhathappenedwhenBrianChesky,JoeGebbia,and Nathan Blecharcyzk tried to pitch Airbnb? Investors like Paul Grahamliterallycouldn’t imaginewhypeoplewouldeveruse theservice.Thisdoesn’thappenbecauseinvestorsaredumb;mostventurecapitalistsandangelinvestorsaresmart,andmostsmart,successfulpeoplewouldprobablyagreethatinvestinginprovenideasisbetterthaninvestinginunprovenones.Theproblemisthat,bydefinition,businessmodelinnovationinvolvestrying
somethingthatisnew,andthusunproven!Inthisbook,we’vetriedtolayoutasetoftools,principles,andpatternsthat
youcanusetodesign,investin,orevaluateaninnovativebusinessmodel.Manyventurecapitalistsliketobragthattheyaremastersof“patternmatching”—butherewemustcautionnotallpatternmatchingishelpful.Thebadkindofpatternmatching is what B-andC-grade investors love—theHollywood high-conceptpitch.ThemovieSpeedwasfamousforitshigh-conceptpitch:“DieHardonabus.”Andifyou’rethefirstpersontomaketheconnection,youmightsucceed.
Speedwasinfactacommercialsuccess,mostlybecauseitdidinfactliveuptoitsdescription.But thesuccessofSpeed led toa raftofderivativeand inferiormovies, ranging fromStevenSeagal’sUnder Siege (“DieHard on a ship”) toStevenSeagal’sExecutiveDecision(“DieHardonaplane”).Whenaninvestorfunds“UberforPets,”that’sbadpatternmatching.The good kind of pattern matching involves understanding what medical
science terms “themechanism of action.” Speed works because confining theactiontoabusthathastostayatacertainspeedorhighertoavoidsettingoffabomb creates built-in dramatic tension—especially given the famously badtraffic in Los Angeles. Airbnb works because it has a large market, becausetravelers spreading awareness from city to city creates virality, and because itfollowstheprovenpatternofanonlinemarketplace.To help you get a feel for applying our principles of business model
innovation,let’spracticebyanalyzingsomeoftoday’sgreatbusinessesandhowtheyfollowthoseprinciples.
ANALYZINGAFEWBILLION-DOLLARBUSINESSMODELS
CASE#1:LINKEDIN
WhenwestartedLinkedInin2002,therecentdot-combusthadledmostpeopleto consider the consumer Internet industry to be dead. The last thing venturecapitalists were willing to do was provide millions of dollars to fund rapidgrowth.Despite this fact, I thought therewas a big opportunity available, andwas able to guide LinkedIn through the start-up growth phase until we couldraisethecapitaltoreallyblitzscale.Thisisthestoryofhowithappened.
MarketSize
The key insight behind LinkedIn was that the Internet was shifting fromanonymouscyberspace toanextensionof the realworld,and thusyouronlineidentitywasanextensionofyourrealidentity.Readersofmygenerationmightremember the famousNew Yorker cartoon with the caption “On the Internet,nobodyknowsyou’readog.”Ididn’tthinkthiskindofanonymitywouldwork
inaprofessionalcontext,hencetheneedforaprofessionalonlineidentity.Andthoughour thesiswas contrarian at the time,my cofounders and Iwere fairlyconfident that the market of “all white-collar professionals” was sufficientlylargetorepresentamajoropportunity.
Distribution
Inorder to raisemoney to scaleLinkedIn,wehad to find away toproveourdistribution strategy. Unfortunately, investors thought of us as “Friendster forbusiness relationships,” which was bad pattern matching and made about asmuchsensetothemas“Tinderforbusinessrelationships”wouldtotoday’sVCs.Instead,wehadtofindawaytousethemoneyandreputationIhadacquiredbyhelpingbuildPayPaltogetLinkedIntothepointwherepeoplewouldinvest.Thefirststepwas toassembleasmall,super-scrappyteam.Wegotourfirst
officeby squatting in thebuildingof a friend’s failing start-up. “Just cleanupafteryourselvessowecangettheleasedepositback,andyoucanuseitforthreemonths,”hetoldme.Ileveragedmyreputationtosecureasmallinvestment,butIknewweneeded toshowsignificantprogress indistributionbeforewecouldraise our next round. Since we didn’t have the capital to pay for traditionalmarketing,weimplementedanumberoftechniquessimilartowhatpeopletodaycall “growth hacking” to get to one million users, which allowed us to raisemoneyfromGreylock.Our core distribution strategy was organic virality, much as it had been at
PayPal.Ouruserswouldinvitetheircontactsviae-mailbecauseithelpedthembuildtheirnetworksandkeeptrackoftheirkeyconnections.Buttheinitiallevelofvirality simplywasn’tenough.Wecouldn’tofferPayPal’skindof financialincentives, so insteadwebuilt things like the e-mail addressbook importer sothatwecould increase thenumberof invitations and let ourusersknowwhentheircontactsjoinedtheservice.
GrossMargins
Grossmarginswereimportantbecauseitbecameapparentthatourusergrowthwas always going to be surpassed by that of the leading consumer socialnetworks. At this point,MySpace had eclipsed Friendster, and Facebook was
quicklygainingonMySpace—andallofthemhadfarmoreusersthanLinkedIn.Our argument was that our professional users were far more valuable, but toprovethatargumentwehadtodemonstrateourabilitytoearnsignificanthigh-marginrevenues.Thefirstbusinessmodelpatternwetriedwasafreemiumsubscriptionservice.
ThefreeLinkedIn.comservicelimitedthenumberofrequestsausercouldsendto friendsof friends (InMails), andwhenusershit those limits, theywouldbeoffered the chance to upgrade to a premium subscription. This subscriptionrevenuewasenough togetus tocash-flowprofitability,but itwasn’tgrowingfastenoughtobetrulycompelling.The key inflection point came when we discovered that companies were
willing to pay for the ability to scan LinkedIn profiles to find the best jobcandidates.Soweofferedittocompaniesasanenterprisesubscriptionproduct,andonceweprovedthatthisnewmodelwasasourceofsignificanthigh-gross-marginrevenues,wehadtheconfidencetoblitzscale.
NetworkEffects
The long-term value of LinkedInwas always intended to come from networkeffects. As a professional social network, LinkedIn leveraged both direct andtwo-sidednetworkeffects,aswellasbecomingastandardformatforpresentingone’sprofessional identity.Thedirectnetworkeffects come from the fact thateachadditionalLinkedInusermakes thenetwork slightlymorevaluable to allotherLinkedInusers.The two-sidednetworkeffectsoccurbecausemoreusersattractmore corporate employers,whilemore employers increase the value ofLinkedInasapassivejob-huntingtool.Finally,bybecominganintegralpartofmost people’s professional online identities, LinkedIn has become a standardthathaslargelyreplacedthetraditionalrésumé.Justoneofthesenetworkeffectswould probably be enough to create first-scaler advantage; all three workingtogetherbuiltamassivestrategicmoatthatprotectedtheLinkedInbusinessfromanynewentrants,andevenfromattemptsbyconsumernetworkslikeFacebooktotakeawaytheprofessionalmarket.
Product/MarketFit
Findingproduct/marketfitforourenterpriseproductwasthekeyinflectionpointin thebusiness.Howdidwedo it?Wefocusedongettingmarket feedbackasquickly as possible.We hired a salesperson, gave him some mock-ups of anenterpriseproduct, and senthim tovisit potential customers. It turnedout thattheyallwantedtobuyit!
OperationalScalability
Blitzscaling LinkedIn presented two major operational scalability challenges,beyondtheobviousoneofsupportingaglobalsocialnetworkwithhundredsofmillions of users. First, to support the business, we actually had to develop,maintain, and update two different products. Without the consumer product,companies wouldn’t see the value of our enterprise product. Without theenterpriseproduct,wecouldn’tmakeenoughmoney tobuildagreatbusiness.We had to do both. It’s hard to find an engineering expert who wouldrecommend fracturing your product and engineering group to work on twolargely separate products, but that’s precisely what we did, despite theinefficiencyandmessiness.Second,wehad to rapidly scale a salesforcewhilewewere still developing
the product they were selling. This took a lot of hard work on the part ofLinkedIn’sCEOs,DanNyeandthenJeffWeiner,andtheirteams.Butwherewecould, we also used technology to help alleviate scaling constraints. Our“Merlin”toolhelpedmakeoursalespeoplemoreproductive(andthusscalable)byautomatingmuchoftheirmanualwork.Merlinwouldanalyzeusagepatternsandtelleachsalespersonwhichcompaniestocall,howtheywerealreadyusingLinkedIn,andevencreateapersonalizedsalesdeckforeachindividualprospect!
CASE#2:AMAZON
MarketSize
Jeff Bezos’s original vision for Amazon was to take advantage of unlimiteddigitalshelfspacetorunastorewhereacustomercouldbuyliterallyanything.Amazonbeganwithbooksbecausethisrepresentedalargeenoughmarketwitha product amenable to e-commerce (durable, fairly standard sizes, readily
available through wholesale distributors). Since then, Amazon has steadilyexpandedfrombooksintomanyotherverticals,andtodayverynearlylivesuptoBezos’svisionofan“everythingstore”(thoughyoustillcan’tbuyautomobilesonAmazon…yet).RetailisatrulygargantuanmarketandAmazonhascapturedan almost unthinkable portionof it and evenmade itsmarketmuchbigger bylaunching AmazonWeb Services. Now, in addition to being “the everythingstore,” Amazon also provides much of the Internet’s computing power,bandwidth,andstorage(includingforotherdominantcompanieslikeNetflix).
Distribution
Amazonwas one of the first companies to fully grasp the possibilities of theInternet as a distribution platform in creating the first successful affiliateprogram, Amazon Associates, which incentivizes individuals and owners ofother websites to refer customers to Amazon in exchange for a share of therevenues generated.This allowsAmazon to turn everyone else’swebsites andonlinecommunicationsintoapowerfuldistributionchannel.Eventoday,ifyousee a book title on the Internet, or in a tweet or an e-mail signature, and youfollow that link, you’ll probably find yourself on Amazon’s website via anaffiliatelink.
GrossMargins
Amazonactuallyscoresfairlypoorlyonthisgrowthfactor,thoughthisislargelya function of the industry rather than being specific to Amazon. Retail is arelatively low-margin business, andAmazon’s devotion to offering low pricesfurther hurts margins. Even today, Amazon’s retail business isn’t profitable(though it probably could be if the health of the company required it; forexample,Amazon’scoreNorthAmericanoperationsareprofitable—it’sjustthatitsprofitsareoutweighedbythelossesgeneratedbyAmazon’seffortsinAsia).YetevenwithinAmazon’sretailbusiness,wedetectsignsthattheselowgross
margins are actually part of a long-term strategy that can generate high grossmargins,evenonretailsales.It’snosecretthatAmazondominatese-commerce;in 2017, analysts like Slice Intelligence reported that Amazon accounted for44percentofUSe-commercesalesin2016,andpredictedthefigurewouldbe
evenhigher in thefuture.Butwhat isoftenoverlookedis thatAmazon’sretailbusiness consists of twovery different units.The first isAmazon’s traditionalretailoperation, inwhichAmazonbuysproductsfromsuppliersandsells themtoitscustomers.Thesecond,farlesswell-knownunitisAmazon’smarketplace,which lets third-party sellers sell their products onAmazon.Those third-partysellersstoretheirinventoryinAmazonwarehousesandpayAmazontodelivertheir products to their customers. If you’ve ever shopped onAmazon, you’veprobably bought a product from a third-party seller; Jeff Bezos has said thatalmost50percentofunitspurchasedonAmazoncomefromthem.Becausethismarketplacebusinessdoesn’trequiretyingupAmazon’scapitalininventory(ittiesupthethird-partysellers’capitalinstead),itsgrossmarginslikelyresemblehigh-margineBay’smorethanitdoeslow-marginWalmart’s.AsBenchmark’sMattCohlernotes,“IsometimeswonderifAmazon’sowned-inventorybusinessisjustamarketinglossleaderandacapital-intensivecompetitivemoat.”WhereAmazon is already tapping into high grossmargins iswith itsAWS
business.Remember, 150 percent of its operatingmargins in 2016 came fromAWS, which accounted for $12.2 billion in revenue and over $3 billion inoperating income. The high gross margins of AWS allow Amazon to investheavilyinmaintainingitsleadoveritscompetitors.Indeed,AWSisestimatedtoholdover40percentofthemarketforcloudcomputinginfrastructure,morethanitsthreebiggestrivals—Microsoft,Google,andIBM—puttogether!
NetworkEffects
Amazonisrelativelyweakonnetworkeffects.Onecustomer’suseofAmazondoesn’tmakeitmorevaluableforanothercustomer,withthepossibleexceptionofAmazon’sproductreviews.Yetwhateverdirectnetworkeffectsexistbecauseof product reviews pales in comparison to the impact of network effects onsomething likeFacebook.Amazon technically isamarketplacewith two-sidednetworkeffects,thankstoitsthird-partysellers,butonesideislargelymissing:Amazon sellers are attracted by Amazon’s massive customer base, butAmazon’s customer base is largely indifferent to those sellers. Amazon doesbenefit from scale effects, and explicitly uses the “flywheel” framework ofauthorandstrategyguruJimCollins.BradStonesummarized thisapproach inhisbookonAmazon,TheEverythingStore:
Lower prices led to more customer visits. More customersincreased the volume of sales and attracted more commission-paying third-party sellers to the site.That allowedAmazon togetmoreoutoffixedcosts like thefulfillmentcentersandtheserversneededtorunthewebsite.Thisgreaterefficiencythenenabledittolowerpricesfurther.Feedanypartofthisflywheel,theyreasoned,anditshouldacceleratetheloop.
YetasimpressiveasAmazon’sflywheelis,whencomparedwiththepowerfulsuperlinear effect of most network effects, it is merely linear or sublinear.Fortunately,Amazondoesbenefitfromstrongnetworkeffectsinoneofitsunits.MostofAmazon’snetworkeffects,likemostofitsgrossmargins,comefrom
its AWS business. The AWS platform benefits from both indirect networkeffects and compatibility and standards. The success of AWS encouragesdevelopers and development products like Docker to rely on it as theirinfrastructure of choice, which makes AWS even more successful (while theemergence of AWS as a standard makes it easier for services built on theplatformtoconnectviaAPI).
Product/MarketFit
Amazonhasrarelystruggledwithproduct/marketfitinitscorebusiness.Forthemostpart,becauseitwastappingintoanexisting—andthriving—retailmarket,Amazonwasabletoleapintohypergrowthalmostimmediately.EvenAWSmetwithrapiduptake,helpedbyAmazon’ssavvydecisiontoleadwithitssimplestproduct, S3 (Simple Storage Service), before expanding to more complicatedones.ItisimportanttorememberthatAmazonhashadmanyfailuresoutsideitscorebusiness.Amazon’spowerfulcore retailoperationsdidn’tallow it to takeoverauctionsorpaymentsfromeBayorPayPal,anditsFirePhonewasacostlyandfruitlessattempttotakeonAppleandAndroid.
OperationalScalability
Amazonhasmanagedoperationalscalabilitysowellthatitmightbethebestintheworldatthistask.
Onthehumanside,JeffBezoshasbeenabletoguideAmazonwithastrongand steadyhandwhile allowingbusiness leaders likeAndy Jassy, theCEOofAWS, or Jeff Wilke, the global head of the consumer business, to run largeportionsofthecompany.ThisdelegationhasallowedAmazontogrowtoover541,900employeesasof2017,makingitoneofthetenlargestemployersintheUnitedStates.On the infrastructure side, Amazon has deftly shifted from minimizing
infrastructurespending,asitdidduringitsearlyyearsbyusingtechniquessuchasoutsourcinglogisticstobookdistributorslikeIngram,tobecomingoneoftheworld’sgreatinfrastructurecompanies.Amazonissogoodatinfrastructurethatits fastest-growing and most profitable business (AWS) is all about allowingother companies to leverageAmazon’s computing infrastructure.Amazon alsomakesmoneybyofferingFulfillmentbyAmazontoothermerchantswhoenvyitsmasteryof logistics,whichought to strike fear into theheartsof frenemieslikeUPS and FedEx. In addition to its eighty-six gigantic fulfillment centers,Amazonalsohasatleastfifty-eightPrimeNowhubsinmajormarkets,allowingit to beat UPS and FedEx on performance by offering same-day delivery ofpurchasesinlessthantwohours.Amazonhasalsobuiltout“sortation”centersthat let it beat UPS and FedEx on price by shipping small packages via theUnited States Postal Service for about $1 rather than paying FedEx or UPSaround$4.50.
CASE#3:GOOGLE
MarketSize
Google’s market size was dramatically underestimated at the outset. WhenGooglecameonthescene,manyconsideredit“yetanothersearchengine”inamarketthatwasalreadydominatedbycompanieslikeYahoo!andLycos.Evenin theunlikelyevent thatGooglewasable tocaptureasignificantshareof thesearchmarket, itwould still be a niche player in comparison to, say,Yahoo!,whichwasaportalwithmajorpropertieslikeYahoo!MailandYahoo!Finance.Observers failed to realize two things. First, Google’s business model
innovation—the relevance-based, revenue-maximizing, self-service advertisingsystemofAdWords—allowedittogeneratefarmorerevenuepersearchthanitspredecessors.Second,theimportanceofsearchwasgrowingatafasterratethan
theInternetasawhole.AstheInternetgrewandtheamountofcontentincreasedat a superlinear rate, so did the difficulty of filtering and finding relevantinformation,makingsearchincreasinglyimportant.CombinethateffectwiththerapidgrowthoftheInternetitself,andtheresultwasamassivemarket.Googlehasastutelyexpandedthemarketsincethenbyleveragingthepower
of its business model to make and monetize key acquisitions like Android,GoogleMaps,andYouTube.
Distribution
Google’stechnologyreceivesmostofthecreditforthecompany’ssuccess,andit is impressive. However, this means that Google’s skillful use of thedistributiongrowthfactorisoftenoverlooked.Togofrom“yetanothersearchengine”to“thelastsearchengine”(asmyold
friend Peter Thiel put it in his 2014 Stanford lecture “Competition Is forLosers”),Googlehadtoleverageaseriesofexistingnetworksandpartners.Forexample,Google’sbolddealtopowerAOL’ssearchresultshelpedthecompanygrow its searchbusiness byorders ofmagnitude.Later, other distributionbetslike theFirefoxpartnership, theacquisitionofAndroid,andthecreationof theChrome browser all paid off and helped maintain Google’s distributiondominance.Googlealsofoundwaystoleveragesmallpartnersaswell,withitsAdSense
program for Web publishers feeding more raw traffic into the AdWordsmachine.
GrossMargins
Google is a phenomenally profitable company, with an enviable margin of61percentin2016.Butthisprofitabilitydidn’thappenbyaccidentorluck;thecredit belongs toGoogle’sAdWords businessmodel. Aswe discussed in oursection on business model patterns, the advertising-supported media modelhasn’t worked for the Internet. Yet when Google first emerged, this was thedominant business model being pursued by major players like Yahoo! andLycos.Google adopted the self-service advertising auctionmodel ofOverture,addeditsownrefinementofselectingadsbasedonconsiderationsofrelevance
and quality as well as bid prices, and pursued a business model of capturingpurchase intent rather than justgatheringeyeballs.Thispurchase intentprovedtobefarmorevaluableperunitoftraffic,enablingGoogletoearnfatmargins.Googlehas sinceused the financial powerof its grossmargins toplacebig
bets that other companiesmight shy away from, such as investing inAndroidand Chrome, two products that were going up against dominant competitors(Apple’s iOS in mobile phone software and Microsoft and Firefox in Webbrowsers).GooglehasalsouseditsmarginstofundradicalexperimentslikeX(formerlyGoogleX)andWaymo(self-drivingcars).Thesebetsmayormaynotpay off, but even if they fail, Google’s margins give it the ability to recoverquicklyandkeepgoing.
NetworkEffects
Google has leveraged network effects quite a bit in its major business lines,thoughnot,ironicallyenough,initscoresearchproduct!Themobile trafficappWaze isaclassicexampleofadirectnetworkeffect.
Wazeharnesseseachuser’s location tocreateamoreaccuratemodelof trafficconditions, while also letting drivers easily report events such as trafficaccidents, speed traps, and stopped cars on the side of the road. ThenWazemakesall thatdatapublic toeveryoneusing theapp. Inotherwords, themoreWazers on the road, the more accurate that road information becomes. Eachadditionalusercreatesvalueforalltheprevioususers.TheAndroidmobileoperatingsystemisaclassicexampleofindirectnetwork
effects.ItsbroadadoptionbyendusersincreasestheincentivesfordeveloperstocreateAndroidversionsoftheirapplications.Theincreasedavailabilityofusefulappsencouragesmorepeopletousedevicesthatrunontheplaftorm.YouTubeisaclassicexampleoftwo-sidednetworkeffects.YouTubebrings
together video creators and consumers—themore content is created, themorepeople show up to consume it. Themore consumers who show up, themoreincentivethereistocreatecontent.Finally, Google’s G Suite provides a great example of the power of
compatibilityandstandards (ironicallyenough,much likeMicrosoftOffice, itsarchrival) aswell as local network effects.When users shareGoogleDocs orGoogle Sheets with others, they lock in anyone who wants to collaborate on
those documents to do the same. This is especially common in individualnetworkslikeaprojectteamorschool.Oncesomeoftheschool’steachersstartusing Google Docs for homework assignments, the pressure builds for all ofthemtostandardizeonGoogleDocs,andforchildrenandparentstoadoptitaswell.Chrisspeaksfromexperiencehere.
Product/MarketFit
Google got the product/market fit for its core search and AdWords productincredibly right. Even from the start,Google’s search resultswere better thanthose of its competitors. But many people don’t realize that it actually tookGooglealongtimetofindtherightproductfortherightmarket.Googlestartedofftryingtosellenterprisesearchappliances,atoolthatsits insideacorporatedata center, indexing content stored on a company’s servers, then offering aGoogle search box to find items within that content. Next, Google tried theadvertising-supported model by running DoubleClick ads; ironically enough,GooglewouldlaterbuyDoubleClick.Fortunately,Googlefoundproduct/marketfitbyrefiningOverture’sadvertisingauctionmodel.Google’sAdWordsproductwas so much better at monetizing search through its self-service, relevance-driven,auctionsystemthatbythetimethosecompetitorsmanagedtoplaycatch-up, Google had amassed the financial resources that allowed it to investwhateverwasnecessarytomaintainproductsuperiority.Google doesn’t always get product/market fit right (and if it had runout of
moneybeforehittinguponAdWords,thesearchbusinessmighthavediedbeforeever achieving that fit). This is a reflection of its very intentional productmanagement philosophy, which relies on bottom-up innovation and a hightoleranceforfailure.Whenitworks,asinGmail,whichwasabottom-upprojectlaunchedbyPaulBuchheit, itcanproducekillerproducts.Butwhen it fails, itresults in killed products, as demonstrated by projects like Buzz, Wave, andGlass. To overcome this risk of failure, Google relies on both its financialstrength (whichcomes from itshighgrossmargins,amongother things)andawillingness to decisively cut its losses. For example, when Google boughtYouTube (which had clearly achieved product/market fit), it was willing toabandon itsownGoogleVideoservice,even though ithad investedheavily inthatproduct.Other massively successful companies take a very different approach. In
contrast toGoogle,wherenewideascancomefromanywhere in thecompanyand there are alwaysmanyparallel projects going on at the same time,Appletakes a top-down approach that puts more wood behind fewer arrows. Applekeeps itsproduct lines smalland tends toworkona singlemajorproductat atime.Onephilosophyisn’tnecessarilybetterthantheother;theimportantthingissimplytofindthatproduct/marketfitquickly,beforeyourcompetitiondoes.
OperationalScalability
Unsurprisingly for an engineering-driven organization, Google excels inoperationalscalability.Foronething,itsheavyinvestmentinitsowntoolsandinfrastructure has allowed its engineering organization to fine-tune itsinfrastructureforhighperformanceasthecompanyhasgrown.Googlehas innovated inpeople scalability aswell.WhilemostofGoogle’s
people management practices are smart but relatively straightforward—forexample,Googleusessmallerteamstoworkonnewproductsandlargerteamsto sustain and grow existing products—Google has invested heavily in peopleanalytics and data to determine things like the optimumnumber of interviewsper candidate (no more than five) and to improve practices for recruitment,performancereviews,andsoon.
CASE#4:FACEBOOK
MarketSize
MarketsizeisoneofthekeyreasonsthatmanyfailedtoappreciatethepotentialvalueofFacebookinitsearlydays.Atthetime,theelevatorpitchforFacebookwouldhavebeen“socialnetworkforcollegestudents.”Thisdescription,whichcombined a new and unproven product categorywith a specific (and narrow)audience,madeFacebooksoundlikeanicheproduct.ButbythetimeIinvestedin Facebook, Mark Zuckerberg’s vision was far broader and more valuable.MarkwantedFacebook tobe thedefaultway thatpeople stayed in touchwiththeirfriends,whichwasandisanenormousmarket.Ofcourse,evenwhenMarkpitchedhisbroadervision,manyinvestorsdidn’tbelievehim,muchtotheirlaterregret.
Distribution
Facebook excelled at distribution.As noted earlier, Facebook’s early focus oncollege students, which caused some to dismiss it as a niche product, wasactually part of an extremely successful distribution strategy. To achieveincredible virality, Facebook would deliberately delay launching at a collegecampusuntilover50percentofthestudentshadrequesteditsothatlocalcriticalmasswasreachedalmostimmediately.Facebook further benefited from leveraging existing friend networks to
expand outward from its original college user base. As users experienced thebenefitsof stayingconnectedviaFacebook, theynaturallywanted toadd theiroff-linefriendstothenetwork.
GrossMargins
LikeGoogle,Facebookstarted its lifewithoutaneffectiverevenuemodel.Butonce it discovered thevalueof sponsoredpostswithin a news feed,Facebookwasabletobecomewildlyprofitable.About90percentofFacebook’srevenuetoday comes from advertising sales, and the company achieves an astounding87percentgrossmargin.ThisgrossmarginallowsFacebooktoinvestheavilyintalentandtechnology.
It has also allowed Mark Zuckerberg to make canny (and expensive)acquisitions, like Instagram and WhatsApp, to become a dominant player inmobile aswell as desktop social networks, and also long-term futurebets likeOculus.
NetworkEffects
We’ve already talked about how Facebook leverages classic direct networkeffects(themoreusersthatjointheplatform,thegreaterthevalueofFacebookto every other Facebook user) and local network effects (once it becomes thedominant social network at a college, it becomes extremely difficult for anyotherplayertopryawayFacebook’susers).Facebookalsoexperiencessomehelpfulindirectnetworkeffectsthankstoits
platformservices,suchastheGraphAPI(whichallowsdeveloperstoleveragethe Facebook social graph of users and their relationships) and Facebook
Connect (whichallowsusers to log in toaWebserviceusingFacebookratherthancreateanewaccountforthatservice).
Product/MarketFit
Facebookachievedproduct/market fit for its core consumer experience almostimmediately,henceitsrapidgrowth.However,partofwhatmakesFacebookagreat company andMark Zuckerberg a great CEO is that Facebook has beenable toachieveproduct/market fit inadditionaland lessobviousareasatotherpointsinthecompany’shistory.ManypeopleforgethowFacebookstruggledwiththetransitionfromdesktop
to mobile. Facebook’s initial mobile product provided a slow, suboptimalexperience,andadoptionof thatproductwasaccordinglyslow.FortunatelyforFacebook,Mark Zuckerberg saw that themarketwas goingmobile and put amoratorium on new feature development in order to focus the entire team onbuildinganew,farsuperiormobileproduct.Inparallel,healsomovedquicklyanddecisivelytoacquireInstagramandWhatsApp;whentheywereannounced,both acquisitions were considered pricey, but in hindsight they were clearlybargains.Today,Facebookhasover1.7billionactivemobileuserseachmonth,and mobile advertising accounts for 81 percent of the company’s advertisingrevenue.Over56percentofFacebookusersaccess the serviceexclusivelyviamobile.EquallyimportantwasFacebook’sabilitytoachieveproduct/marketfitforits
advertisers. When Facebook began, the conventional wisdom was that user-generatedcontentlikeFacebookwouldneverbeabletoattractadvertisers,whowouldnotwant theirbrandsappearingwithpoor-qualityoreven inappropriatecontent. Google’s search model was what worked in online advertising.Facebook was able to overturn the conventional wisdom by developingalgorithms to block inappropriate content, and by learning from Twitter’ssponsored updatemodel and incorporating ads into the FacebookNews Feed.Thenewsfeedmodelhasbeenespeciallyeffectiveformonetizingmobileusage.Inareturntowhatworkedintheprintworld,advertisementsareintermixedwithcontent,andasyoupage through themagazineor scroll through the feed,youencounter advertisements as part of your normal flow, as opposed to theinterruptionsofpop-uportakeoverads,ortheeasilyignoredstaticplacementofthe traditional banner ad. Yet Facebook’s News Feed is even better for
advertisers than amagazine, because Facebook’s core social actions (clicking,liking,sharing) trainusers toengagewithwhateverappears in theNewsFeed,includingadvertisements!
OperationalScalability
How did Facebook successfully overcome the growth limiter of operationalscalability? On the technology side, one of the philosophies that helpedFacebook become successful was its famous motto “Move fast and breakthings.”This emphasis on speed,which camedirectly fromMarkZuckerberg,allowedFacebooktoachieverapidproductdevelopmentandcontinuousproductimprovement.Eventoday,everynewsoftwareengineerwhojoinsFacebookisasked to make a revision to the Facebook codebase (potentially affectingmillionsorevenbillionsofusers)onhisorherfirstdayofwork.However,asFacebook’suserbaseandengineeringteamgrewtoamassivesize,Markhadtochange the philosophy to “Move fast and break things with stableinfrastructure.”While this new motto might seem self-contradictory, Mark explains that it
focusesonahigher-levelgoal.“Thegoalistomovefast,”Marktoldme.“Whenweweresmaller,beingwillingtobreakthingsallowedustomovefaster.Butaswe grew, the willingness to break things actually started slowing us down,becauseincreasingcomplexitymadeitharderandhardertofixthingsoncetheybroke.Bytakingtheextratimetofocusonstableinfrastructure,wereducetheimpactandtimetorecoverfrombreakingthings,sothatwecanactuallymovefaster.”
WHATCOMESAFTERASTRONG,PROVENBUSINESSMODEL?
If you believe you’ve designed a business model that can support massivegrowth and value creation, the next step is to decide on your strategy. That’swherestrategyinnovationcomesin.
PARTIII
StrategyInnovation
Whileblitzscalingis themaintopicof thisbookandthesecretweaponbehindthestaggeringgrowthandmarketdominationofhundredsof theworld’smostvaluable companies, it is also a strategy innovation. It is in fact the strategyinnovation that supports itsownecosystemof rapidgrowth in the faceof riskand uncertainty. To blitzscale or not to blitzscale is a strategic (and difficult)choice, andbecauseof thiswewant to takea lookatwhenandhowfoundersandCEOsapproachthatdecision,andhowitchangestheircompaniesandeventheirownrolesintheirbusinesses.
WHENSHOULDISTARTTOBLITZSCALE?
Hereisthequestionwearemostoftenaskedbyfoundersofstart-upswhenwetalkaboutblitzscaling:WhenshouldIstarttoblitzscalemycompany?One of the reasons you might find it challenging to grasp and apply the
principlesofblitzscaling, especially ifyou’reanexperiencedexecutive, is thatdoing so requires you to throw out many of the normal rules of business. Itbasically takes everything you thought you knew from years of hard-wonexperienceorfrombusinessschoolorfromobsessingaboutstayingleanduringyour early start-up phase…and hurls it out the window. Careful planning,cautiousinvestment,courteousservice,andatightlycontrolled“burnrate”(theamount of cash the company consumes eachmonth tomake payroll, pay therent,andsoon)mayendupbeing tossedaside in favorof rapidguesstimates,ignoringangrycustomers,andinefficientcapitalexpenditures.Whywouldyoueverwant to pursue such a risky and unintuitive course of action? In aword,
speed.Remember, the objective of blitzscaling is to achieve “lightning” growth
despite the increased risks and costs. The only time that it makes sense toblitzscale is when (whether for offensive or defensive reasons) you havedeterminedthatspeedintothemarketisthecriticalstrategytoachievemassiveoutcomes.You don’t necessarily need to have solved your revenue model before
deciding to blitzscale. In fact, a key element of blitzscaling is often thewillingness of investors to fund growth before the revenuemodel is proven—afterall,it’sprettyeasytofundgrowthaftertherevenuemodelisproven.Slackhadspentnearlyfiveyearsand$17millionondevelopmentpriortoits
publiclaunchinFebruary2014.Justtwomonthslater,beforetheendofApril,ithad raised another $43 million. Both of these investments took place beforeSlack had proven its revenue model and started generating significant sales.Slack’sfreemiumbusinessmodel(offeringafreeserviceandencouragingusersto upgrade later to becoming paying customers) meant that even after twomonths of rapid user growth, the company hadn’t proven its ability to makemoney.FortunatelyforSlackanditsinvestors,thisaggressivenesspaidoff.Astheinitialwaveoffreeusersstartedconvertingtopaid,Slackwasabletoraiseanadditional$120millionsixmonthslatertoaccelerateitsgrowthevenfurther.Every $100 billion scale-up blitzscaled to get there, but that doesn’t mean
everystart-upcanorshouldblitzscale.Ifyourproduct/marketfit isn’tright,oryourbusinessmodeldoesn’tworkyet,orifthemarketconditionsaren’trightforhypergrowth,thenprematureblitzscalingcanleadohsopainfully(andrapidly!)to“blitzfailing.”Sadly, premature blitzscaling can sometimes kill a nascent market by
“poisoningthewell”sodramaticallythat investorsandentrepreneursavoidthespace. For example,Webvan’s notorious failure kept most players out of thegrocerydeliveryspaceforoveradecade.Herearea few factors to look for ifyouarewonderingwhether the time is
rightforyourcompanytoblitzscale.
ABIGNEWOPPORTUNITY
To achieve massive success, you need to have a big new opportunity—one
wherethemarketsizeandgrossmarginsintersect tocreateenormouspotentialvalue, and there isn’t a dominant market leader or oligopoly. A big newopportunityoftenarisesbecauseatechnologicalinnovationcreatesanewmarketorscramblesanexistingone.ShishirMehrotra, the formergeneralmanagerofYouTube, visited our Blitzscaling class at Stanford and explained howtechnologicalchangescreatedabignewopportunityforYouTubetoexploit:
Why was YouTube at the right time? Networks were finally bigenough to streamvideo.Cell phone cameras allowed everyone torecord videos. And the investment environment allowed a verycapital-intensivebet.
Ifthegrossmarginsofthisnewopportunityarelow,themarketsizehastobeevenbiggertomakeitabigopportunity.Youhavetoknowthattheultimatesizeoftheprizeisworthit.Thecostofblitzscaling,evenwhensuccessful,isusuallyquitehigh.Itsimply
isn’tworth the risk andpain touseblitzscaling topursue a small opportunity.The good news is that in the Networked Age, the ability to rapidly expandproductsandservices intoa trulyglobalmarketmeans that therearemorebigopportunitiesthaneverbefore.Consider the rise of Alibaba. Jack Ma realized that the opportunity for e-
commerce in China and other Asian markets was an even bigger long-termopportunitythane-commerceintheUSmarket.WhenJackfoundedAlibabain1999, the e-commerce market in China was negligible and lacked keycomplementary resources like the equivalent of FedEx, UPS, Visa, andMastercard(andPayPal).Yetheknewthattheultimateprizewasasbigastheycome.TheOrganization forEconomicCooperationandDevelopment (OECD)has predicted that China’s middle class (defined as a household income ofbetween$20,000and$160,000peryear)willreach73percentofitspopulationby2030,makingitsmarketsizenearlytripletheentirepopulationoftheUnitedStates.Suchaprizejustifiesanextremelyhighlevelofinvestment.Jackraised$25millionfromSoftBank,GoldmanSachs,andFidelitytogrowthebusiness,andanother$75millioningrowthequityfromGeneralAtlanticin2009.Today,Alibaba controls an estimated 80 percent of the e-commercemarket in China(the same figure forAmazon in theUnitedStates is 44percent), and its 2014IPO on the NewYork Stock Exchange became the largest in history, raising
$25 billion for the company. In July 2017, Alibaba became the first Asiancompanytosurpass$400billioninmarketvalue.Some big opportunities are so enormous that they spawn secondary
opportunities for blitzscaling. For example, Alibaba’s Taobao Marketplacesupportscountlessmerchants,Facebook’srisecreatedtheplatformforZynga’sinitial growth, and Apple’s iOS devices created a big opportunity for gamedeveloperslikeRovioandSupercell.
FIRST-SCALERADVANTAGE
Themostfrequentoffensivereasonforblitzscalingistoachieveacriticalmassthatconfersalastingcompetitiveadvantage.Sometimesthisissimplyamatterof capturing economies of scale, aswithAmazon orWalmart, butmost oftencriticalmasstriggersnetworkeffects,aswithUberorAirbnb.Blitzscaling is unlikely to prove successful if another company has already
achieved first-scaler advantage. During the dot-com era, both Amazon andYahoo!attempted frontal assaultsoneBay’s auctionbusiness,but thenetworkeffectsofeBay’stwo-sidedmarketplaceofbuyersandsellersmeantthatitsfirst-scaleradvantagewastoostrongtoovercome.Incontrast,whenAmazonenteredthe business of sellingmusicCDs—yes, once upon a timemusicwas sold onphysical disks—which lacked network effects, it quickly destroyed theincumbentmarketleader,CDNow.First-scaler advantage can also be specific to a particular market or set of
customers. Latin American e-commerce giant MercadoLibre was founded in1999,whenAmazonwasalreadygeneratingbillionsinrevenue,andeBaywasalready aggressively expanding overseas. Yet despite not being the global e-commerce first scaler,MercadoLibrewasstillable tobuildavitalbusinessbybeing the first scaler in LatinAmerica. In an interview for Reid’sMasters ofScale podcast,MercadoLibre’s founder andCEO,MarcosGalperin, explainedwhyhewasabletoachievefirst-scaleradvantage:
BeforeIstartedMercadoLibre,IactuallydidasurveywithtwentyLatin American students that were colleagues of mine at theStanfordGraduateSchoolofBusiness,andtheyallsaidthatthis[aneBay forLatinAmerica]would neverwork inLatinAmerica.Atthattime,eBaywasbasicallysuccessfulandoperationalintheUS,
Germany,andJapan.
ByjumpingintoamarketwhereevenotherLatinAmericanentrepreneursfearedto tread,MercadoLibre was able to gain a head start on the competition andachievefirst-scaleradvantage.It’s importantnot toconfusecriticalmasswith first-moveradvantage.Being
first to launch in amarketmight earn you congratulations on being a productvisionary,butifyouaren’talsothefirsttoscale,you’llendupasafootnoteinaWikipediaarticleaboutyourcompetitorwhodid.Furthermore, sometimes there is no first-scaler advantage to bewon. If you
can’tidentifyanynetworkeffectsorcustomerlock-in,scalingmightnotconfersufficient advantage to warrant blitzscaling. For example, we suspect that themarketforfooddeliveryfromexistingrestaurants—apurecommoditybusiness—isunlikely tooffer any lasting competitive advantages thatwould justify anexpensiveblitzscalingcampaign.
LEARNINGCURVE
Anotherwaytouseblitzscalingtocreatealastingcompetitiveadvantageistobethefirsttoclimbasteeplearningcurve.Someopportunities,suchasself-drivingcars,requireyoutosolvehard,complexproblems.Themorerapidlyyouscale,the more data you have to drive learning (or train machine learning), whichimprovesyourproduct,makingiteasiertoscalefurtherinthemarketwhileyourcompetitorswhohavejustbeguntolearnlagfarbehind.Netflixistheleaderinstreamingvideoentertainment,butitonlyachievedthis
statusbybeingwillingtoclimbaseriesofsteeplearningcurves.Rememberthesituation Reed Hastings faced when he started Netflix in 1997: the dial-upmodems that connected most consumers to the Internet were far too slow tostream high-quality video content. So Netflix decided to compete with videostores like Blockbuster by offering a subscription service (with no hated latefees!)tomailmovieDVDstoconsumers’homes.ThismeantthatNetflixhadtoclimb a steep learning curve in terms of both DVD-specific tasks, such asnegotiating with the studios for access to movie DVDs and coordinating thelogistics required to ship them to and from consumers, and developing newfeaturesliketheabilitytorecommendmoviesbasedonpastselections.Climbingthelearningcurveforthesetaskswaspainfulandexpensive,butitgaveNetflix
acompetitiveadvantageoveritscompetitors.Later, as broadband connections became more widespread, Netflix had to
climbthelearningcurvewhenbuildingoutitsmassivestreaminginfrastructurewhile continuing to improve its consumer recommendation engine. That waswhenNetflix began running into amajor strategic issue.Netflix relied on thestudiosfor itscontent(moviesandTVshows),but thestudiosnowsawonlinevideocompanieslikeYouTubeandNetflixasathreat.Inresponse,theybegantoincreasethepricetheydemandedfromNetflixforlicensingtheircontentandheld back some of their “crown jewels” (e.g., massively popular content likeSaturdayNightLive)forthemselvesandHulu(anindustryjointventure).The logicalconclusionwasclearbutdaunting.Netflixneeded todevelop its
own original content. Now the company had to climb what was perhaps itssteepestlearningcurveyet,sinceitwouldbecompetingwithHollywoodstudiosthathadnearlyacenturyofexperienceintheirfield.NetflixhiredTedSarandosasitsheadofcontent,andsuccessfullyclimbedthislearningcurve,justasithadclimbedsomanyothersinthepast.Today,Netflixmightverywellbetheleaderinoriginalvideocontent,andeventraditionalHollywoodpowerplayers,suchassuperproducer Shonda Rhimes (Grey’s Anatomy, Scandal, How to Get AwaywithMurder)andcomedianAdamSandler(HappyGilmore,GrownUps),haveswitched from traditional studios to Netflix. What’s more, the other learningcurvesthatNetflixclimbedalongthewayactuallyhelpeditbeat thestudiosattheir own game. The consumer recommendation engine gives Netflix anunprecedented ability to predict what content its users want to watch, whichallows it to work with creators to produce that content (such as the populardramaStrangerThings).AndbecauseNetflixhasgreaterconfidenceinitsownpredictions than its competitors have in theirs, it can outbid them for contentwhentheygohead-to-head.
COMPETITION
Yet despite these offensive reasons to scale, the most common driver ofblitzscaling is the threat of competition.Evenwithout competition, youwouldstillwanttoachievefirst-scaleradvantageandclimbthelearningcurve,butyoumightprefer the less risky fastscaling approach togrowth.Askyourself, “Cansomebodyelserealizethisopportunitybeforeme?”Iftheanswerisyes,movingfaster probably reduces the risk of competitionmore than it raises the risk of
failure.Themoreintensethecompetition,thefasteryoushouldtrytomove.RememberthesituationthatfacedBrianCheskyandAirbnbinthespringof
2011? Just as the business began to take off, the company faced a terrifyingcompetitor in the form of the Samwer brothers of Germany and their rapidlygrowing European Airbnb clone, Wimdu. Chesky and his cofounders wereforcedtomakeaharddecision:sticktobusinessasusualinSanFranciscoandriskbeingtrouncedbyWimdu…orblitzscaleandwin.Lookingbackafewyearslater,Cheskyadmittedthatthecompetitionforcedhishandforthebetter.TheAirbnb/WimdustoryisbecomingmorecommonintheNetworkedAge.
The world used to have a lot more businesses that were protected fromcompetition by geographic fragmentation—such as regional newspapers andphysicalbookstores—muchlikeDarwin’sfinchesontheGalápagosIslands.Theriseofboth theInternetand theNetworkedAgehasconnected those“islands”into a single, hypercompetitive market, with fierce competition for a fewdisproportionately valuable leadership positions. Because person-to-personinformation exchange occurs so quickly and seamlessly today, ourcommunications networks have accelerated the process by which individualmarketpreferencesresultindominantsuppliers.Today,webuyourbooksfromAmazon,anditsfounder,JeffBezos,ownstheWashingtonPost.Oneofthereasonsbusinessestendtorelyonblitzscalingisthatspeedisone
oftheprimaryadvantagestheyholdvis-à-vislargecompanies.Start-upscanactquicklytocapitalizeonthenewopportunitiescreatedbytechnologicaladvances.Iftheydawdleandproceedatthesamepaceasabigcompany,they’refightingon an even playing field, whichmeans that the big company’s resourceswilllikelyconfermassiveadvantage.
GOODTIMES,BADTIMES
While it may seem like blitzscaling is a strategy that only works in “hot”markets, it can be successful under anymarket conditions. The key nuance isthatacompany’srateofgrowthneedstobemeasuredonarelativeratherthanabsolutescale.Inarapidlygrowingmarket,acompanythatgrows100percentper yearmight be losing share; during turbulent times, a company that grows50 percent per year might be gaining enough share to achieve marketdominance. You can successfully blitzscale in good times, and you cansuccessfully blitzscale in bad times, thoughmarket conditions can and should
affectyourstrategy.Hotmarketsmakeiteasiertoattractthecapitalandtalent(especiallycapital)
toplowintoblitzscaling.Uber isaclearexampleofhowaccess tocapitalcanfund aggressive and inefficient growth that may confer long-term strategicbenefits.Uber’sabilitytoraisebillionsofdollarshasallowedittosubsidizeitsservicetoattractmoredriversandpassengers,reinforcingthenetworkeffectsofits two-sided marketplace. Plentiful capital has also allowed it to expandaggressively intoothermarkets inanattempt tobeat itscompetition tocriticalscale.Evenafterascandal-plagued2017,UberstilldwarfsitsUSarchrivalLyft.In July 2017,Lyft announced that it had reached onemillion rides per day, amilestonethatUberachievedattheendof2014.Duringthedismaldaysofthedot-combust,Googlefollowedtheblitzscaling
playbook by using a distribution deal with AOL to dramatically expand itsAdWords business. The deal, first announced in May 2002, gave AOL an85percentshareoftherevenuegeneratedbyAOLsearchespoweredbyGoogle,withaguaranteedminimumof$150millionperyear.Atthetime,Googlehadlessthanone-tenththatamountinthebank.Thismayhaveseemedrisky,giventhattheNASDAQhadfallennearly80percentfromitshightwoyearsearlier,butitispreciselythisperceivedriskthatprobablyallowedGoogletooutbidtheincumbentproviders, thepublicly-tradedOvertureandInktomi.Yetwhileboththe revenue share and guarantee were highly aggressive, Google’s improvedAdWords algorithmsmade the deal highly profitable for both parties, and themove allowed Google to increase its revenues from roughly $19million pre-AOLin2001to$347millionpost-AOLin2003,anearlytwentyfoldjump.Noonetrulyknowswhetherthemarketswillgoupordowninanyparticular
year. But regardless of which direction they move, blitzscaling can be a keystrategyforcapitalizingonthebiggestopportunities.
GOINGFASTER
Onceyoudecide toblitzscale, thekeyquestionyouneed toaskandanswer is“How can wemove faster?” This isn’t simply a matter of working harder orsmarterwiththesameresources.It’sdoingthingsthatothercompaniesnormallydon’t do, or choosing not to do things that they do because you’rewilling totolerategreateruncertaintyorlesserefficiency.For example, in 2015, PayalKadakia, the founder of ClassPass (amonthly
subscription service for fitness classes) decided that she needed to double thesizeofherstaffinjustthreemonthssothatClassPasswouldbeableexpandintomore cities. To achieve this kind of speed, Kadakia and her team abandonedtraditional hiring processes and followed two simple rules. First, they hiredpeoplefromtheirpersonalnetworks,withanemphasison“branded”talent.Forexample, if an employee had a friend, and that friend worked for themanagement consulting firm Bain & Company, that friend got hired becauseClassPass could assume that the person was smart and would get along withpeople.Second, someof the time savedbynot interviewing for skills allowedthe team to interview for alignment with the company’s mission. Crazy?Perhaps.ButClassPasswas inacrowded,emergingmarket, andbeingable tohire faster than the competition helped it maintain and increase its leadershipposition.Blitzscaling also requires a strong focus on risk management. While
blitzscaling requires risk taking, it doesn’t require unnecessary risk taking.Indeed, the higher level of risk associated with blitzscaling makes riskmanagement even more valuable and important. As Yahoo! cofounder JerryYang told us in an interview for Reid’sMasters of Scale podcast, “All boldstrategieshavearisk.Ifyoudon’tseeit,you’reflyingrisk-blind.”A finalword of caution—just because youcan blitzscale doesn’tmean that
you should. Throwing out the rules of business doesn’t guarantee success anymorethanfollowingtherulesdoes.IntheearlydaysofLinkedIn,weknewthatachievingacriticalmassofusers
wasgoingtobeachallenge.Wehadtodoalotofeducationtogetprofessionalsto understand our value proposition. Most didn’t realize the power of theirnetworks, andhow technology couldhelp themenhance, extend, and leveragethembetter.One approach,whichquite a number of people recommendedwefollow, was to raise a large amount of venture capital and embark on anaggressive advertising campaign to accelerate user growth. This would be aclassic example of blitzscaling—sacrificing efficiency for growth against abackdropofuncertainty.Butwedecidedagainst thisstrategy;webelievedthatthe competitionwasn’t as urgent asmany thought, and keeping our burn ratelowerwouldallowustowaitforthemarkettocatchuptoourpointofview.Aswe pursued our “slow and steady” start-up growth strategy, the people whorecommended investing in inefficient growth warned us that our competitorswouldleaveusbehind.Weweren’tworried,sinceourreadingofthemarketwasthatcompetitors likePlaxodidn’t trulyunderstand thepowerofaprofessional
social network (instead, theywere treating their product like an addressbook)and thus weren’t competing for the same market. This hypothesis ended upbeingprovenbylaterevents.If taking on additional cost and uncertainty doesn’t actually confer an
advantage, it’sbetter tofollowthe traditional rulesofbusiness(at least for thetime being) so that when blitzscaling does become appropriate, yourorganizationcanbeefficient,wellmaintained, andmore ready to scale.WhenLinkedInfinallydididentifytheopportunitytogrowamajorbusinessbysellinganenterpriseproducttorecruiters,wewereamorematurecompanythatcouldblitzscalewithconfidenceinourabilitytomakethesejudgmentcalls.
WHENSHOULDISTOPBLITZSCALING?
Whileblitzscalingisapowerfulstrategy,itisnotapermanentone.Nobusinesscan grow forever, simply because no market is infinite. You blitzscale whenyour market is big or growing fast—or preferably both. If your market stopsgrowingorreachesitsupperlimit,youshouldstopblitzscaling.Because blitzscaling is—by definition—an inefficient use of capital, it only
makessensewhenspeedandmomentumare important.Blitzscaling is like theafterburnersonafighterjetthatallowyoutoflyatdoubleortriplenormalspeedbut consume fuel at a shockingly high rate. You don’t just switch on theafterburnersandneverturnthemoff.Oneofthemajorchallengesofblitzscalingisknowingwhenyourbusinessis
outgrowing your current strategy, and when you need to change course. It’sunwisetowaituntilyoustopgrowingtomakethetransition.Instead,youshouldpayattentiontosomeoftheleadingindicatorsthatcanactasanearly-warningsignthatyou’veoutgrownyourstrategy:
Decliningrateofgrowth(relativetothemarketandcompetition)
Worseninguniteconomics
Decreasingper-employeeproductivity
Increasingmanagementoverhead
Whentheseleadingindicatorsbegintoappear,itisprobablyasignthatyour
current strategywon’t scale further,and it’s time tobegin thecycleanew.Forexample,Yahoo!wasable to ride its core strategyofbeing the leadingonlinemedia company for a decade, with revenues growing briskly (albeit with adowndraft during the dot-com bust) through 2005. At that point, however,Yahoo!’srevenuesstoppedgrowing(andindeedbegantodeclinein2007,evenbefore the global recession hit). Google had just passed Yahoo! in annualrevenues in 2005 ($6.1 billion forGoogle, $5.3 billion forYahoo!), and afterthat the fates of the companies diverged dramatically. Yahoo!’s revenue wasroughlyflatin2006,whileGoogle’snearlydoubledagain.Blitzscaling can actually be dangerous when you reach the limits of your
market. Ifyourunoutofmarketheadroom,all thatspeedandmomentumwillcometoacrashinghaltasyouslamintoyourmarket’sceiling.Theusualsymptomofrunningoutofheadroom,besidesasuddenslowdown
ingrowth,isinternalconflict.Managersandinvestorswhohavegottenusedtocontinuinggrowthstartaskingquestionslike“Whatwentwrong?”and“Who’sresponsible?” If thecompanydoesn’t realize the rootcause, themostcommon(andunhelpful)responseistocallforchangingtheCEOortheexecutiveteam—theVPofsalesisparticularlyvulnerablebecauseheorsheoftentakestheblamefor theslowdown—orboth.Howmany timesdoes replacing theCEOactuallyreignitemassivegrowth?TheonlygoodexamplewecanthinkofiswhatSteveJobsdidatApple.So ifyouhaveaSteveJobswaiting in thewings,goaheadandswitchCEOs.Otherwise,itprobablywon’thelp.Consider what happened to two blitzscalers who ran out of headroom—
GrouponandTwitter.Grouponwasoneofthefastest-growingcompaniesofalltime,thankstoitsleadershippositionintherapidly-emergingdailydealsmarket.Unfortunately, the daily dealsmarket suddenly stoppedgrowing.Theproblemwas actually an ironic echo of bad blitzscaling—Groupon merchants usedGroupon’s daily deals as an inefficientway to generate rapid revenue growth,onlytodiscoverthatthepromotionsdidn’tleadtorepeatbusinessoranyotherlong-termcompetitiveadvantageorvalue.Groupon began suffering from internal turmoil, and, sure enough, Andrew
MasonwasreplacedasCEO.Itdidn’thelp.What Groupon should have done is to stop blitzscaling. The pursuit of
inefficient growthwas overheating themarket andmaking it unsustainable. IfGrouponhadreducedthediscountsitrequiredofmerchants,growthwouldhavedeclined,butthebusinessgeneratedbythosesmallerdiscountswouldhavebeen
moresustainable.Twitterranintoasimilarproblem.Attheendof2014,itsusergrowthslowed
to a crawl. This was the signal for Twitter to ease off the gas and focus onefficiency.Intheperiodfrom2011to2014,Twitterhadincreaseditsemployeebase by more than a factor of ten, anticipating continued growth. Twittercontinued to hire in 2015, adding nearly three hundred employees despite thelack of user growth. Twitter may have been “head faked” because revenuecontinued to grow as the advertising market matured. Revenues more thandoubledoverthecourseof2015,thenstoppedgrowing.Twitter is beginning to shrink its number of employees today, but probably
shouldhavebeenevenmoreaggressiveindoingsoonceitbecameapparentthatitsblitzscalingperiodwasover.Naturally, it was during this period that Twitter decided to change CEOs
again,withDickCostolo(whohadreplacedfounderEvWilliams)departingandJackDorsey assuming the role of interimCEO.BothCostolo andDorsey areincrediblytalentedexecutives,butevengreattalentcan’tblitzscaleinamarketthathashititsceiling.
CANICHOOSENOTTOBLITZSCALE?
First,aswe’vediscussedabove,blitzscalingisnotforeveryone.Forexample,in1994, thesameyear thatJeffBezosfoundedAmazon, therestaurateurThomasKellerboughtTheFrenchLaundry inYountville,California,andturnedit intooneoftheworld’sgreatestrestaurants,winningacovetedthree-starratingfromtheMichelin Guide. Today, Amazon has over 541,900 employees and is themarket leader in online retail, ebooks, cloud computing, andmore,while TheFrenchLaundry,withlessthanfiftystaffmembers,inasinglelocation,servingjustsixtycustomersperday,isstilloneoftheworld’smostfamousrestaurants.BothAmazonandTheFrenchLaundryaregreatbusinesses,buttheyexistin
fundamentallydifferentworlds.Amazon’sbusinessreliesonmassivescaleandbillions of dollars of infrastructure; The French Laundry relies on localingredientsofthehighestquality,preparedbysomeofthemostskilledcooksinthe world. Scale is critical to e-commerce and cloud computing; scale isantitheticaltoworld-classfinedining.ItisasimpossibletoimagineAmazonasa small, independent bookstore as it is to imagine The French Laundry as aglobalrestaurantchain,vyingwithMcDonald’sforfranchisesupremacy.
Butiftheconditionsareripeforblitzscaling,competitorsmaychoosetotakeontherisksyou’rereluctanttoassumeinexchangeforachanceatreapingthepotentialrewards.That’swhatAirbnblearnedwhenWimduentereditsmarket.Blitzscaling requirescapital—whether from investorsor fromcash flow—to
fund relatively inefficient growth. If investors are willing to act quickly andprovidelargeamountsofcapital,theriskthatacompetitordecidestoblitzscaleishigher.Thesameistruewhenabusinessmodelprovidesalotofhigh-marginrevenue to fundgrowth.So the safest time tochoosenot toblitzscale iswhenyou’re pursuing a relatively low-margin business model that investors areunwillingtofundatall,unwillingtofundatscale,orunwillingtofundquickly—like,say,afine-diningrestaurant.Manysmallor“lifestyle”businessesfallintothiscategory,whichmakestheir
decision to avoid blitzscaling perfectly rational.However,markets can changequickly. Let’s return to 1994 and the founding of Amazon. For many years,independentbookstoreshadcarvedoutamarketnichebypositioningthemselvesrelativetocompetitionfromchainstoressuchasBarnes&NobleandBorders.The rise of Amazon and its pursuit of a blitzscaling strategy significantlychanged the competitive landscape for those bookstores, forcing them torespond. In 1994, the American Booksellers Association had over 8,000members;by2009,thatnumberhaddeclinedto1,651,downnearly80percent.Shockingly, that number has grown every year since 2009, rebounding to
2,321 in2017.We’lldelvemore intohowindependentbookstoresmanaged tosurvive in the ageofAmazonwhenweexaminehow todefendyourbusinessagainstblitzscalingcompetitors.Evenifyoudon’tfacesuchacompetitor,blitzscalingstillmakesmajorwaves
that can impact your business.Here in SiliconValley, blitzscaling has causedhigherpropertyvalues,ahighercostofliving,andatightlabormarketthathasaffectednearlyeverybusinesswithinitsboundaries,regardlessofindustry.Evenifyoudon’tcompetewithblitzscalersforcustomers,youprobablycompetewiththemforofficespaceandemployees.
BLITZSCALINGISITERATIVE
Successfulblitzscalingisanexerciseinserialproblemsolving.Eachofthefivestagesrequiresdifferentsolutionstothesamebasicproblemsofpeople,product,finance, and soon.Each timeyoumanage to solve aproblem, theproblem is
never solved forever, it’s only solved for now. As the company continues togrow,youhave to solve the sameproblemagain,under anewandpotentiallyradicallydifferentsetofcircumstances.In2013,PaulGraham,thecofounderofYCombinator,wroteafamousessay
titled“DoThingsThatDon’tScale,” inwhichheargues thatstart-upsare likeold-fashioned cars with engine cranks. To get them started, founders need toengageinaseparateandlaboriousprocessthatcouldn’tpossiblyworkatscale,suchaspersonallyrecruitingaproduct’sfirstusers.Thisessayisaclassic,butitmaygivesomereadersthemistakenimpressionthatoncethe“engine”starts,allyouneedtodoiskeepdoingthingsthatscale.Inotherwords,theconventional(anderroneous)wisdomsays:
Step1:Dothingsthatdon’tscale.Step2:Achievescale.Step3:Dothingsthatscale.
Butwhenyou’re blitzscaling, the things youdo to help you scale up to thenextstageprobablywon’tallowyoutoscaleuptothestageafterthat.Tobuildatrue scale-up, almost everything you do has to change with each stage.Blitzscaling extends the simple three-step process of “Do Things That Don’tScale”asfollows:
Step1:Dothingsthatdon’tscale.Step2:Reachthenextstageofblitzscaling.Step3:Figureouthowtodoonesetofthingsthatscale,whilesomehow
alsofindingawaytodoacompletelydifferentsetofthingsthatdon’tscale.
Step4:Reachthenextstageofblitzscaling.Step5:Repeatoverandoveruntilyoureachcompletemarketdominance.
Thisdoesn’tmeanthatyoudon’tneedtoplanahead.Althoughyouwilloftenneed to do things that don’t scale, at the same time you will have to makechoices that permit the possibility (though not the certainty) ofmassive scale.For example, if your core businessmodel lacks scale advantages and networkeffects, and the only possible go-to-market strategy consists of door-to-door
sales, it’s unlikely that you’ll ever be able to build a massively importantbusiness,whetherornotyouattempttoblitzscale.
HOWBLITZSCALINGSTRATEGYCHANGESINEACHSTAGE
Aswesawinourdiscussionofblitzscalingindifferenteconomicenvironments,speedisalwaysrelative.Whatrepresentshypergrowthspeedatonestagemightbeonlyaverageduringanother.Forexample,nearlyeverystart-uptriestomovequickly.ThismeansthatduringtheFamilyandTribestagesofblitzscaling(upto one hundred employees), it can be challenging to move at a speed that isclearlyfasterthantheaveragestart-up.Thereareonlythreewaystoaccomplishthis.First,youmightbe theonlycompetentplayer inyourmarket space.This is
extremely rare, because any attractive market space tends to draw in smart,aggressiveentrepreneurs.Second, youmight be the first player in yourmarket to have figured out a
brilliant growth strategy (wonderful if you can manage it but also somewhatrare).Forexample,PayPalwasn’ttheonlypaymentsstart-uptryingtoattackthemarket, but itwas the first to tap intoviralmarketing for extremely rapid andcost-effectiveuseracquisition.Third,youcandistinguishyourself fromyourpeersbypursuing scalemore
resolutely. Start-ups that assume success and make commitments andinvestments accordingly can get a jump on their rivals—provided the marketplays out as they anticipate. This kind of confidencemanifests itself in beingmore aggressive about fund-raising, hiring, and infrastructure investment—incurring current expenses that hopefully will allow your company to movemuch faster in the future. Throughout its history, Amazon has been moreaggressivethanitscompetitors,andthataggressivenesshaspaidhugedividends.Of course, it helps that Jeff Bezos and his team are world-class at executingagainstthisstrategy.Thedownside,ofcourse,isthatthecostoffailureismuchhigherthanifyou
proceeded with deliberate caution and waited for proof before makingcommitments.But thisadditionalcostcanbedwarfedby thepotentialbenefitsof achieving first-scaler advantage in a valuable winner-take-most or winner-take-allmarket.At theVillage (hundreds of employees) andCity (thousands of employees)
stages,thespeedsofcompetingorganizationsbecomemuchmorevaried.Somewill be content with focusing on optimizing for efficiency (scale-up growth),whileotherswillfocusonspeed(fastscaling)orspeedinthefaceofuncertainty(blitzscaling).At this stage,blitzscaling is lessabout rawaggressionandmoreaboutpursuingadifferentiated(butstillaggressive)strategy.Forexample,oneof thesignaturestrategiesforblitzscaling is rapid,parallel
market development.WhenAirbnbmade the decision to blitzscale, its chosenstrategy was to rapidly expand from a single office in the United States to ascoreofofficesaroundtheworld,especially inEurope.Thiskindofgrowthishighly inefficient—think of all the new knowledge and infrastructure andpersonnelanorganizationhastoacquiretosuccessfullyopenofficesaroundtheworld—but itcanallowacompanytostandoutfromitscompetitors. ItwouldhavebeenmoreefficientforAirbnbtoexpandonecountryandoneofficeatatime,refiningitsapproachbasedonthe lessonsofeachrollout,but thatwouldhavealloweditscompetitorWimdutobethefastermover.Inotherwords,whenitneededtogrowfromaforty-personcompanytoaglobalcompanyinasingleyear, Airbnb couldn’t afford to be cautious with its capital and focus onefficiency.We’ll see this pattern of simultaneousmarket development in laterexamplesandacrossmultipleindustries.AttheNationstage(tensofthousandsofemployees),thestrategyshiftsagain.
Companies get tomassive scale by dominating an industry until they becomematureandmainstream.AsdescribedinGeoffreyMoore’sCrossingtheChasm,Nation-stage companies have succeeded in crossing the chasm between acustomerbaseofearlyadoptersand“MainStreet.”Marketdominancemakesitdifficult to grow much faster than the overall market, while market maturityreduces thenumberofopportunities fororganicgrowth.Asa result, scalingatthisstageisaboutincubatingandgrowingamajornewbusiness.In 2007, Apple had over twenty thousand employees, was the dominant
company in the online music business, and was a successful player in thepersonal computer business. Meanwhile, Google had over ten thousandemployees and was the dominant company in search. Nokia, the dominantmobile phone handset maker, had over seventy thousand employees. TheseNation-stagecompanieswereallstartingoutwithroughlythesameshareofthemarketforthenewfangled“smartphones.”In 2007, Apple introduced the iPhone and Google introduced the Android
operating system.Three years later, they dominated themobile phonemarket,
whileNokiawasindisarray,eventuallysellingitshandsetbusinesstoMicrosoftin 2013. The ability to blitzscale new businesses in a new market is whatseparatedAppleandGooglefromNokia,andit’swhathaspoweredtheirascenttothenumberoneandnumbertwomostvaluablecompaniesintheworld(asof2017).
HOWTHEROLEOFTHEFOUNDERCHANGESINEACHSTAGE
Theroleafounderplaysintheblitzscalingprocesschangesineachstage(andanemployee’s role relative to the founder will likely also change). As theorganizationgrows,thespecificskillsrequiredtoleaditevolveaswell.
Stage1(Family):TheFounderPersonallyPullstheLeversofHypergrowth
Intheveryearlydaysofacompany,afounderhastodoeverything,includingimplementingthetechniquesofblitzscaling.Forexample,ifyourbusinessreliesonviralmarketingfordistribution,youprobablywilldoeverythingfromwritingthe copy for the invitation e-mails to segmenting the data on open andconversionrates.
Stage2(Tribe):TheFounderManagesthePeopleWhoArePullingtheLevers
As the organization grows, a founder probably starts to manage a team ofemployees.Evenifyouretainsomeindividualresponsibilities,thebulkofyourvaluecreationcomesfromworkingwiththeteammembersandhelpingthembemoreproductive.Forexample,ifyounowruntheengineeringteam,youmightstillbedoingsomeworktomaintaintheearliercodeyouwrote,butyourfocusshouldbeonmanagingtheotherengineersandlettingthembuildnewfeatures.
Stage3(Village):TheFounderDesignsanOrganizationThatPullstheLevers
TheVillage-stagetransitioncanbedifficultforyouasafounder,becauseitisat
this phase that it becomes harder to see the immediate impact of your work.Whileyoumightknowandinteractwithfrontlineemployees,you’renotlikelyto be their directmanager anymore.Nowyouneed to take a big-picture viewand focus on designing the organization. Founders who don’t find thisinteresting or appealingmay choose to remain individual contributors or teammanagers.Thisisalsothestageatwhichorganizationshireexecutivesfromtheoutside;we’lldiscussthisfurtherinthenextpart.
Stage4(City):TheFounderMakesHigh-LevelDecisionsAboutGoalsandStrategies
WhenthecompanyreachestheCitystage,thefounder’sroleistomakethebigstrategicdecisions.Thesedecisionsmayverywellhavetacticalimplications,butnow it’s someone else’s job to work those out. At Facebook, one of the keyhigh-level decisions that Mark Zuckerberg made was to halt new featuredevelopmentfornearlytwoyearstofocusonFacebook’smobileproduct.Whenhe made this gutsy decision in early 2012, Facebook was deep into the Citystage, with over four thousand employees. He didn’t personally hire thedeveloperswho joined themobile team,ordesign thenewmobile app,buthemade the tough call, then held accountable thosewhowere pulling the leversdirectly.
Stage5(Nation):TheFounderFiguresOutHowtoPulltheOrganizationBackfromBlitzscalingandStartBlitzscalingNewProductLinesandBusinessUnits
Even though managing a Nation-stage company has some things in commonwithmanaging a traditional business, it’s critical to keep blitzscaling, even asyou implement some traditional management practices. When Steve Jobsreturned to Apple, for example, he both focused on traditional measures ofoperationaleffectivenessandinvestedinbuildingnew,insanelygreatproducts.On the traditional management side, he slashed inventories and improvedApple’s financialmanagement,buthealso launchedmajornewproducts, suchastheiPod,iTunes,theiPhone,andtheiPad.
FROMSTRATEGYTOMANAGEMENT
Whenacompanyisblitzscaling,theconstantdoublingoreventriplingofitssizemakes it difficult to apply traditional management techniques designed forenvironments inwhich15percent representsbrisk annualgrowth.As a result,successfulblitzscalershavetoimplementmanagementinnovationstosteertheirburgeoning organizations through their growing pains. The upcoming sectionswilldiscusshow.
PARTIV
ManagementInnovation
Oneof thekeyfeatures thatsetsglobalgiantsapartfromthosecompanies thatflameout or implodebefore they can reachmarket dominance is an ability toevolve and optimize theirmanagement practices at each stage of growth. Theproventechniqueswe’lldescribeinthispartfallintotwomaincategories:eightkeytransitions,whichhelpguidethecompanythroughthestagesofblitzscaling,and nine counterintuitive rules, which turn the conventional wisdom oftraditionalmanagementonitsheadinordertocopewithblitzscaling’sfrenziedpaceofgrowth.Whetheryouareatthehelmofacompany,runaspecificdivision,orleada
smallerteam,eachofthefollowingtechniquescanofferyouguidanceonhowtomanagegrowthasyourorganizationprogressesfromstart-uptoscale-up.
EIGHTKEYTRANSITIONS
TRANSITION#1:SMALLTEAMSTOLARGETEAMS
Thefirstandmostobviousmanagementchallengeforblitzscalingorganizationsto navigate is the shift from small teams to large teams. Even if a rapidlygrowingcompany tries toorganize itself asacollectionof small teams, it stillrequires averydifferent approach topursue its corporategoals and initiatives.Nor is growth simply a matter of turning a crank. Every aspect of peoplemanagement,fromrecruitingtocoachingtocommunications,hastoadapttothedifferentstagesofblitzscaling.
Smallteams,whichareespeciallycommonintheFamilyandTribestagesofblitzscaling, can operate spontaneously and informally thanks to the personalrelationships and frequent contact between team members. This flexibilityallowstheseteamstobeextremelyadaptableandtochangedirectionsquicklyasthecompanylearnsnewinformationandhastoadjustitsstrategyandtactics.During theFamilyandTribe stagesofPayPal, havinga small, nimble team
allowed us to execute four hard pivots during the first year of the company’sexistence.When Peter Thiel, Max Levchin, and Luke Nosek founded PayPal(then known asConfinity) inDecember 1998,Confinitywas intended to be amobile phone encryption company using Max’s highly efficient encryptiontechnology.Fromthere, thecompanypivotedfirst tomobilephonecash(pivot#1) and then to PalmPilot payments via infrared beaming (pivot #2).Unfortunately, the network of PalmPilot userswhowanted to beammoney toeach other simply wasn’t that robust, so we pivoted again and added e-mailpayments (pivot #3). By the end of the year, we saw an emergingmarket insettlingeBaytransactionsandpivotedourproductdevelopmenteffortstoservethatmarket(pivot#4).In just twelve months, we had launched a company, built a product, and
pivoted four times! This was possible only because between eight and fortypeopleworkedforthecompanywheneachofthepivotstookplace,allowingustoeasilyshifttheentirebusinessfocusandtacticsquicklywitheachpivot.As the business grows into the Village stage and beyond, its organization
necessarily includes larger teams, such as departmentswith tens of employeesoften dispersed across various offices and places. These larger teams cannotoperate spontaneously and informally; an individual employeemight only seecertainteammembersafewtimesayear,ifever.Coordinatingtheeffortsoftensorhundredsofindividuals—andensuringalignmentwiththegoalsoftheentireorganization as awhole—requires planning and formal processes, often to thechagrin of an idealistic founder more interested in long-term vision than theminutiaeofday-to-daymanagement.WendyKopp,thefounderofTeachforAmerica,learnedthislessonthehard
way.InaninterviewforReid’sMastersofScalepodcast,shetoldus,“Istartedouttwenty-eightyearsagowithatotaldisdainfororganizationalmatters.Ijustthought everyonewho comes to this should bemission-driven, andwe’re notgoing tohaveanyhierarchy,andwe’regoing topayeveryone thesame thing.Aboutfiveyears into this, I realizedif Ididn’tbecomeobsessedwith thevery
mundanematterofhowtomanageeffectively,we’dnevergetthere!”Yetbeyondsimpleorganizational logistics,oneof themajorchallenges that
leadersofblitzscalingorganizationsneedtoovercomeisthepsychologicaleffectthatthistransitionhasonearlyemployeesandevenfounders.At the Family stage, it’s often the case that every member of the team is
involvedineverymajordecision.AttheVillagestageandbeyond,thisisnearlyimpossible.Employeesarebusyenoughsimplykeepingupwiththeactivitiesoftheir immediate teamorarea; theoperationsofotherdepartmentsare largelyamystery.Fornewemployees,thisstateofaffairswillseemnormal,butforearlyemployees,thisshiftcanbedisconcerting,leavingthemfeelingliketheyusedtobeinsidersbutarenowtreatedlikeoutsiders.Theanswerisnottoinvolvethoseemployees in every decision—that would be inappropriate and logisticallyimpossible. Rather, create other systems to help them feel connected to thecompany’smission.Forexample,TheAllianceoutlineshowtoursofdutykeepemployeesengaged.Visitalliedtalent.comformoreinformationandresources.You’re also adding qualitatively different kinds of people to the team.One
metaphorIusetoexplainthisshiftistotakeyetanotheranalogyfrommilitaryhistory: themarines take thebeach, thearmy takes thecountry,and thepolicegovern the country.Marines are start-up peoplewho are used to dealingwithchaosandimprovisingsolutionsonthespot.Armysoldiersarescale-uppeople,whoknowhowtorapidlyseizeandsecureterritoryonceyourforcesmakeitoffthebeach.Andpoliceofficersarestabilitypeople,whosejobistosustainratherthandisrupt.Themarinesandthearmycanusuallyworktogether,andthearmyandthepolicecanusuallyworktogether,but themarinesandthepolicerarelyworkwell together.As you blitzscale, youmay need to find new beaches foryourmarinestotakeratherthanaskthemtohelppatroltheexistingones.The expansion of the organization may also create issues around career
expectations.One of the topicswe’ll discuss later is the need to swap in keyexecutivesasthecompanygrows.Mostpeoplehaveskillsandexperiencesthatoptimizethemforaparticularstage,andnoteverypersoncaneffectivelygrowinlockstepwiththecompany.SimonRothmansawthisfirsthandwhenhewashelpingeBaytoblitzscale.“Peoplehaveelasticlimits,”hetoldus.“Ofthefirsthundredpeople,onlyafewscaledtoaten-thousand-personorganization.Itwashard to predict who would scale. People who were smarter than me didn’talwaysscale.”Here in SiliconValley, it’s quite common for an executive to specialize in
getting a company from zero to $1 million, and for another to specialize ingettingitfrom$1millionto$10million.Thiscancausefrustrationamongearlyemployees, especially if they were leading a specific function and now anoutside executive is being hired above them to be their boss. That’s why it’simportant tosetcorrectexpectationsupfront.Beclear thatemployeeswillgetopportunities to grow and advance their careers, but this doesn’t necessarilymeanthatifthey’rerunningengineeringnow,they’llbeVPofengineeringwhenthe company has ten thousand employees and is planning its IPO. Focus onresponsibilityinsteadofthespecifictitle.Anemployeewhorunstheengineering“department” at the Family stage might consider it a demotion to be one ofseveraldirectorsofengineeringattheCityorNationstage,butyoucanpointoutthatattheFamilystageshewasmanagingateamofthreeengineersandnowsheoverseesateamofonehundred.Encourageemployeestofocuslessontheirjobtitles andmore on how each tour of duty’s activities and experiences preparethemforgreaterresponsibilitiesinthefuture.Thisfacetofthetransitiontolargerteamscanbethemostdifficulttomanage,
butitiscrucialtoblitzscalingsuccess.Noonelikesfiringemployeeswhohavebeentherefromthebeginning,butthinkofitthisway:ifyourexecutivescan’tscale, your business won’t scale either. The ideal solution is to retain earlyemployeesinnewrolesthatadvancetheircareersandhelpthecompany,butifyouhave tochoosebetween losingacherishedemployeeandallowinghim toflounder in a role for which he isn’t suited, it is better to have an honestconversationandanamicablepartingthanitistoallowboththeemployeeandultimatelythecompanytofail.
TRANSITION#2:GENERALISTSTOSPECIALISTS
Another importantorganizational arc ismoving fromgeneralists to specialists.Duringtheearlystagesofblitzscaling,theneedforspeedandadaptabilityplacesaheftypremiumonhiringsmartgeneralistswhocangetmanydifferent thingsdone in an uncertain and rapidly changing environment. But as the companygrows, it needs to shift to hiring specialists who are less fungible but haveexpertiseinanareathatiscrucialtoscalingtheorganization.Thisisn’ttosaythereisnoplaceforgeneralistsatblitzscalingorganizations.
Infact,oneofthemainbenefitsofbringingonspecialistsisthatitallowsyoutoredeploycapablegeneraliststoattackyourmostpressingchallenges.
For example, when LinkedIn was still in the Tribe stage, one of the earlyemployeesIhiredwasMattCohler.FreshoutofMcKinsey,Mattwasabrilliantyoungmanwhowantedtogetintothestart-upworld.Iexplicitlyhiredhimasageneralistand,oncehewasonboard,usedhimasafirefightertotacklethemosturgentproblems.Atthetime,ourbiggestareaofneedwasrecruiting,soMatt’sfirst jobwas to lead the recruiting function. Itwasn’t something that his prioreducationorworkexperiencehadpreparedhimfor,butIknewhewassmartandscrappy, and I counted on him figuring it out.He performed admirably in therole,thenmovedontotackleotherfires,bothformeandforMarkZuckerbergatFacebooklateron.(Today,MattisageneralpartneratBenchmark,aventurecapitalfirm.)Google even codified the value of generalists in its Associate Product
Manager (APM) program, an initiative Marissa Mayer founded because shebelievedthathiringtechnicalpeoplestraightoutofcollegeasproductgeneralistswould result in flexible, adaptable employees who could fill a lot of needs.Today, distinguished APM alumni include Quip founder/CEO (and formerFacebook CTO) Bret Taylor, Asana cofounder Justin Rosenstein, andOptimizelycofoundersDanSirokerandPeteKoomen.Specialistsalsoplayakeyrole.ConsiderPatWadors,LinkedIn’sformerchief
human resources officer. Pat joined us in 2013, during the City stage of ourgrowth,andhelped leadus intoNationhood(she recently leftLinkedIn to joinmy friend John Donahoe, former eBay CEO, at ServiceNow—a return to theCitystage).LikeMatt,Patisbrilliantandtalented,butalsoaspecialistwhohasheld HR roles at leading companies such as Viacom, Merck, Yahoo!, andPlantronics. Running a major function for a City-or Nation-stage companyrequiresdeepdomainexpertiseand isn’t something thata smartgeneralist canjust“figureout”inamatterofweeks.Whilehiringspecialistsisanincrediblypowerfultooltohelpyouscale,itis
dangerous to do so prematurely. Specialists are just that—specialized. Whilethey are talented enough to be able to do things outside their specialty,redeployingthemseldomallowsthemtogenerateanywherenearthevaluetheycanprovidewithintheirspecialty.Forexample,IhavenodoubtthatPatissmartenough to learn JavaScript programming, but Iwould question thewisdomofasking her to give up her role to go through a coding boot camp and joinServiceNow’sengineeringteamasanentry-levelsoftwareengineer.ItwouldbeacriminalwasteofPat’stalentandlikelyabadfinancialmoveforthecompany.
Inaddition,thetransitionthatoccurswhenyoubringinspecialiststomanageor replace generalists can strain themorale of the organization. “Demands forfunctionalexpertiseoftenoutstripearlyemployees’abilitiestokeepupthroughorganic learning,” wrote Harvard Business School’s Ranjay Gulati and AliciaDeSantola in “Start-Ups That Last: How to Scale Your Business,” whichappeared in the March 2016 issue of the Harvard Business Review. “As aconsequence, functional leadership titles increasingly go to outsiders, and thelegacy folksmay grow resentful. Early employeesmay also chafe against thenarrowing confines of their changing roles. Not every generalist can or evenwantstobecomeaspecialist.Oftenpeoplegetfrustratedandleave,takingtheirvaluable relationships and their tacit understanding of the firm’s mission andculturewiththem.”In many cases, you should work to retain your generalists, both for their
culturalandinstitutionalknowledge,andfortheirabilitytotacklenewproblems.But if you are unable to do so, and early generalists decide to leave theorganization, you should try to maintain a positive relationship with them asmembersofyourcorporatealumninetwork.Youcanreadmuchmoreaboutthistopicinourpreviousbook,TheAlliance.IntheFamilystage,youshouldhireonlygeneralists.Youcancertainlyfollow
thetraditionalapproachofrecruitingfromelitecollegesorhiringex-McKinseyanalysts, but you should also try to focus on people with prior experience atearly-stage start-ups who list a broad range of responsibilities andaccomplishments.Theymaynothavethepedigree,buttheyaregreatatlearningnewthingsandatcharginghardtoexecuteonthem.Plus,theearlybusinessisintoo much flux to effectively leverage the finely tuned capabilities of a truespecialist. Even at the Tribe stage, hiring a specialist should be considered amajorexception—forexample,ifyouneedanengineerwithaveryspecializedareaofexpertise,suchasdatascienceormachinelearning.TheVillagestageiswhere it becomes prudent to hire specialists, as both executives and keycontributors. At the Tribe stage you want employees with skill sets flexibleenoughtopivotalongwiththecompany,butifyouhavehundredsofemployees,you better have some pretty well-developed theories about your business andwhereitisgoing!AlmostanyexecutivehireattheCityorNationstageisgoingtobeaspecialist.Butevenattheselargest,lateststages,youshouldstillmixinsomenumberofgeneralists.Thinkofgeneralistsasthe“stemcells”ofyourorganization.Yourbodyhasa
smallnumberofstemcellsthathavethecapabilitytomorphintovariousother
typesofcellsasneeded.Inalargeorganization,youmayneedasmallnumberofpeoplewhocanperformvariousfunctionsasneeded,whetherexploringnewproductsandtechnologiesortacklingissuesthatlackawell-definedsolution.
TRANSITION#3:CONTRIBUTORSTOMANAGERSTOEXECUTIVES
The terms “manager” and “executive” are often used interchangeably. Webelieve that managers and executives play very different roles. Most of theconfusion probably comes from the tendency in early-stage start-ups for thesamepersontoplaytheroleofmanagerandexecutive.Theseareseparateroles,evenwhenthesamepersonplaysthem.Managers are frontline leaders who worry about day-to-day tactics: they
create, implement, and execute detailed plans that allow the organization toeitherdonewthingsordoexistingthingsmoreefficiently.Bycontrast, theroleoftheexecutiveis toleadmanagers.Forthemostpart,
executives don’tmanage individual contributors. Instead, they focus on visionandstrategy.Yettheyarestillconnectedtothefrontlineemployeesbecausetheyarealsoresponsibleforthe“fightingspirit”oftheirorganizations;theyneedtoberolemodelswhohelppeoplepersistthroughinevitableadversity.Both executives andmanagers are necessary for successful blitzscaling, but
theyplaydifferent roles at different stages.Whena company is in theFamilystage, itmay not need any formalmanagers.And even if it does, that role isgenerallyfilledbythefounder/CEO.AsthecompanygrowsintoaTribe,itwillneedmanagers to run the various functional departments, such as engineeringandsales.Thesemanagersmaybe the founders,or theymaybeoutsidehires.Theirkeyobjectiveistomakeasmallteamproductiveonaday-to-daybasis.When the company reaches the Village stage, it will need executives. It
simply isn’t possible to coordinate a company with hundreds of employeeswithout executives to manage and lead multiple managers. Let’s imagine acompanywithsixdepartments:engineering,sales,marketing,product,support,and administration. If each department head managed ten direct reports, andeachmanagerreporteddirectlytotheCEO,themaximumnumberofemployeesunder this executive-less arrangementwouldbe sixty-seven (eleven in eachofsix departments, plus a founder/CEO). This is still small enough as to bemanageable, but were the company to grow beyond that, it would becomeimportanttobuildinanexecutivelayertokeepthingsrunningsmoothly.
A company in the Village stage might have hundreds of employees. Theengineering department alone would require multiple teams and team leaderswhowould report toaVPofengineeringchargedwithcoordinating the teamsandarchitectingtheoverallorganizationofthedepartment.One of the typical challenges we discussed in the section on the transition
fromsmallteamstolargeteamsistheneedtorecruitexecutivesfromoutsidetheorganization. This represents a major change in approach for a company thatprobably promoted fromwithin to this point, rewarding early employeeswhoemergedasnaturalleaders.However,thetransitionfrommanagertoexecutiveisgenerallyfarmoredifficult in theseorganizations than that fromcontributor tomanager. Every employee has likely reported tomanagerswith varying stylesand qualities;when promoted to a first-timemanager, they can drawon theseexperiences to help develop their own management style. But when anorganization needs executives for the first time, internally promotedmanagerscan’tdrawontheexperienceworkingwithexecutivesatthatcompany—becausethereweren’tany.Therearenorolemodelstoprovideguidance.We call this situation the “Standard Start-up Leadership Vacuum,” and the
resultisthatinexperiencedfoundersfindthemselveshavingtohireandintegrateexperienced executives from the outside. The situation is made worse whenthosefounderswaituntil thestrainon theorganizationhasbecomeunbearablebefore making the new hires, meaning that all the leaders are new to thecompanyprecisely at the timewhen tension anduncertainty are runninghigh.The key to navigating this transition is open-mindedness: insiders need to beopentotheoutsideideasof thenewexecutives,while theoutsidersneedtobeopentolearningfromwhathappenedbeforetheyarrived.Nooneisbornanexecutive,andveryfewmakethetransitionfrommanager
toexecutivewithoutstumblingalongtheway.Hiringoutsideexecutivesletsyouoffload that often painful and expensive education to those executives’ prioremployers.However,ablitzscalingorganizationcan’t simplyhireanyonewithexecutive experience at another companyof similar or slightly larger size.Anexecutiveatalargercompanymaynothaveanyexperiencewithblitzscalingoreven start-ups. Running a hundred-person department at a hundred-year-oldcompanythatgrowsat5percentayeardoespracticallynothingtoprepareyoufor running a hundred-person department at a company that is tripling in sizeeveryyear!Atthesametime,youdon’twanttohiresomeonewhoseexperiencewithblitzscalingisatacompanymuchlargerthanyours.Aswe’lldiscussalittlelater,insteadofhiringfortheskillsyouthinkyoumightneedinthefuture,you
shouldbehiringfortheskillsyouneedrightnow.The ideal, of course, is to hire an executive with past experience at a
blitzscaling start-up that has already dealt with the challenges your companycurrently faces. This is why investors have more confidence in serialentrepreneurs. One of the major advantages that companies in Silicon Valleyenjoy is generations of rapidly scaling companies that have produced a richsupplyofexecutiveswithblitzscalingexperience.Yetevenifyoucan’tlandanideal candidate, second best is to hire amanager who has previously workedwithsuccessfulexecutivesinaveryrapidlygrowingcompany,oranexecutivewhoearnedherexecutiveexperienceatalargerormoretraditionalbusinessbutwhoalsoworkedatablitzscalingstart-upatanothertimeinhercareer.Consider the case of Facebook.Mark Zuckerberg hired Sheryl Sandberg in
partbecauseshehadexperienceblitzscalingasanexecutive,havinghelpedhergroup within Google grow from a handful of people to over four thousandemployees.AndoneofthekeythingsSheryldidthathelpedFacebookscaleupto theVillage, City, andNation stageswas to fill critical leadership positionswithother experienced scale executives, such asMike “Schrep”Schroepfer asVP of engineering and David Ebersman as CFO. Schrep had learned how toscale engineeringorganizationsatMozilla,whereheoversawmassivegrowth,and had also founded his own start-up, CenterRun, before that. David hadpreviouslyworked asCFO of the biotech leaderGenentech and had firsthandexperience with the rapid growth associated with blockbuster drugs such asHerceptinandAvastin.MartinLauplayedasimilarroleforPonyMa(MaHuateng)andtherestofthe
foundingteamatTencent.Maandhiscofoundersweresmarttechnologistsbutlackedbusinessexperience,especiallyoutsideChina.Lauhad the internationalbusinessexperiencefromhisworkwithGoldmanSachsbut,crucially,alsohadastrong engineering background and could relate to the team. Lauwas able tobring much-needed organizational best practices to Tencent, such as revenuegoalsandlong-termplans.“Thiswasadisciplinethatwasurgentlyneededforayoung company growing extremely fast,” said Hans Tung, a partner at theventure capital firm GGV Capital, who coinvested with Tencent in DidiChuxing.Anotherhelpfulstrategyforhiringoutsideexecutivesistobestrategicabout
howyoublendthoseoutsidehireswithinsidepromotions.MariamNaficyoftheonline art and graphic design marketplace Minted realized that she could
combine the strengths of both groups to create a more effective managementteam. “It takes years and years to grow candidates fromwithin,” she told ourBlitzscalingclassatStanford.“Wetakedisciplineswherewearen’tstrong,likefinance and HR, and hire in experts from the outside.When it comes to oursecretsauce, likecrowdsourcing,wegrowpeoplefromthe inside.OurVPArtandStationerygrewinternally,whileourVPFinanceandChiefPeopleOfficerareoutsidehires.”Evenifanoutsideexecutivehastherequisiteblitzscalingexperience,though,
heorshecouldstillfailbecauseofpoorculturalfit—the“transplantrejection”phenomenon.Whenhiringanexecutivefromanothercompany,therearethingsyoucanandshouldtakeintoaccounttohelpensurethegraft“takes”tothehostculture.OnemasterofthesetechniquesisJohnLilly,aventurecapitalistatGreylock
Partners and the formerCEOofMozilla.WhenhewasCEOatMozilla, Johnoversawincrediblyrapidgrowth;duringhisfirstsixmonthsattheorganization,thenumberofemployeestripled.GivenMozilla’ssmallinitialsize,thisgrowthnecessitated hiring executives from the outside (“the graft”), which wasparticularly challenging because of the company’s strong engineering-drivenculture(“thehost”)thatwasalreadyskepticalaboutoutsiders.Johnwasabletodo this successfullyby following the same three-stepprocess thatwasused tohirehim.
1. Hiresomeonewho isalreadyaknownquantity toat leastonememberofthe team. John was hired byMitchell Baker, his predecessor as CEO ofMozilla. The two had gotten to know each other by serving on a boardtogether,andMitchell’spersonalendorsementofJohncarriedweightwiththe team at Mozilla. Similarly, John had known Schrep at Stanford andworkedwithhimatJohn’sownstart-up,Reactivity,beforehehiredSchrepatMozilla.
2. Bring the new executive in at a lower level initially and let the executiveprovehimselforherself. Johngavehimself the title“DirectorofBusinessDevelopment andOperations” and only took on bigger titles after he haddemonstratedhisabilityandvaluetoexistingteams.Heemployedthesametechnique when he hired Schrep, bringing him in as “Director ofEngineering.”OnceSchrephadachance toprovehimself,Johnnoted,“It
becamepretty clear to everyoneprettyquickly thatSchrepwas incrediblyconfident,andimprovedeverythinghetouched.”ThisvisiblesuccessmadepromotinghimtoVPofengineeringbothobviousanduncontroversial.
3. Once the executive has earned the team’s trust and credibility, considerpromotinghimorher.AnotherexecutivethatJohnhired,DanPortillo,wasbroughtintorunrecruitingbutprovedsovaluablethathewaspromotedtoVPofpeopleandaskedtorunHRaswell.Today,DanservesinasimilarroleatGreylock.
AsyourcompanyprogressesfromaVillagetoaCityorevenaNation,you’llcontinuetoneedtohireexecutives,bothbecausethegrowthinsizewillrequireyou toadd layersaboveyour frontlinemanagers, andbecauseyourexecutiveswon’t always have what it takes to scale to the next stage. But once yourorganization has successful executives who can serve as role models andmentors,youwillbeabletostartpromotingpromisingmanagerswithpersonalexperience working with those successful executives from within. WhenFacebook was growing, it was critically important to bring in experiencedexecutives like Sheryl Sandberg, but almost all of Facebook’s key productleaderstodayweretrainedinternally.While entrepreneurs often resist creating a hierarchy by classifying their
peopleintoexecutives,managers,andcontributors,thiskindofformalstructureis essential to growth, according to RanjayGulati andAliciaDeSantola, whowroteintheHarvardBusinessReviewin2016:
When launching their start-ups, many founders eschew hierarchybecause of their egalitarian ideals. But as their firms scale, agrowingnumberofpeoplereporttoahandfulofleaders.Foundersmay think this allows them to remain in command, because alldecisionspassthroughthem.Butironically,theirorganizationsspinout of control as centralized authority becomes a bottleneck thathindersinformationflow,decisionmaking,andexecution.Acoupleof people at the top can’t effectively supervise everyone’sincreasingly specialized day-to-day work; in such a system,accountabilityfororganizationalgoalsgetslost.
GulatiandDeSantolacitetheexampleofCloudflare,whosefounderspubliclycommitted to building a completely flat organizationwithout hierarchy or jobtitles.Whilethefoundersmadethisdecisionforacommendablereason—CEOMatthew Prince felt that eliminating job titleswould prevent early employeesfromfeeling“demoted” if thecompany laterhiredmoreexperiencedpeople—the results, as documented in a Harvard Business School case study by TomEisenmann and Alex Godden, were poor: “In the three months ending inJuly2012,fiveofthefirm’sthirty-fiveemployeesquit,somecitingthelackofaclear midlevel reporting structure and the nonexistent HR practices. Theydescribedsituations inwhichtheyhadnooneto turnto(shortofpesteringthefounders)iftheythoughtcertainpractices,suchasactivitiesrelatedtosoftwareorcodingstandards,neededtochange.”Blitzscalingorganizationsneedorganization,notjusttocoordinatetheirmany
resources and activities, but in order to maximize speed. The organization’scollective learning rate—especially within its leadership team—determines itsability to anticipate future trends,while the strength of its internal structure—especiallyintermsofitsfrontlineteams—determinesitsabilitytoactquicklyonthosekeyinsightsandseizethecompetitiveadvantage.
TRANSITION#4:DIALOGUETOBROADCASTING
One area that undergoes the most change during blitzscaling is the internalcommunications process. As the company grows, you have to shift frominformal, in-person, individual conversations to formal, electronic, “push”broadcastingandonline“pull”resources.Youalsohavetoshiftfromsharingallinformationbydefaulttodecidingonwhatissecretandwhatisshareable.Ifyoudon’t manage to develop an effective internal communications strategy, yourorganizationwillbecomedisjointedandstarttofallapart.In theFamilystage, theentireorganization is typicallyunder thesameroof,
possibly even allworking in the same room.As a result, information spreadsquite naturallywithout any additional intervention—possiblymore than you’dlike.Whenyouhaveaquestionorneedfeedback,youcansimplypopupfromyour chair (or balancing ball or treadmill desk) and say, “Hey! Does anyoneknow…?”This “prairie dog” style of communication is organic, quick, and effective.
Everyone is still working on the same initiative, so the interruption is likely
relevant and/or productive (or easily ignored by wearing headphones ifnecessary).Thebiggestchallengeyoulikelyfaceatthisstageiskeepingtherarevirtual employee in the loop.Because it is so easy for the rest of the team tocommunicate with one another, you have to make an effort to constantlycommunicatewithanyremotemembers tokeep themon thesamepageas therest of the team.Communications tools likeSlacknot onlyprovide amediumwhere all teammates participate on the same terms but also allow forasynchronous communication, which helps overcome time zone differences.Another approach some companies take is to set up a 24/7 videoconferenceusingtoolslikeSkypeorGoogleHangoutstosimulatebeinginthesameroom.These informal bonds remain a critical part of the communications process,
evenasyour firmgrows intoaglobalgiant.Humanbeingsaresocialanimals,andthebondsbetweencoworkersandteammatesrequireregulardialogue.However, as early as the Tribe stage, youwill need to begin implementing
processes to supplement the one-to-one dialogue. For example, nearly everyTribe-sizestart-upholdsaweeklycompanymeeting,thoughwithwildlyvaryingdegreesofeffectiveness.Theweeklymeetingismosteffectivewhenitservesasamechanismtobringtogethertheentirecompany,andforcompanyleaderstoconveykeymessagestoemployeeswithwhomtheydon’tworkdirectly.A Tribal meeting should be well organized, with an agenda and other
materialsprovided inadvance so thatparticipants canengage inan interactivediscussion rather than simply listen to senior leaders talking or, worse, sufferthrough text-dense PowerPoint presentations. The goal shouldn’t be to makedecisionsinthesemeetings(unlessthetopicisoneonwhicheveryonecanandshouldhaveinput,likewheretoholdtheholidayparty);rather,it’stomaximizeinput fromsmartpeopleandmakesure thateveryonefeelsheard.Asa leader,youshouldseekoutopinionsfromacrosstheorganizationonimportantissues,butyoucan’tabdicateyourresponsibilityandrelyonlyongroupconsensus tomaketoughdecisions.The best Tribal meetings include rituals that go beyond the cut-and-dried
businessofthecompanyandhelptheemployeesgettoknowoneanotherbetteraspeople,notjustworkers.Forexample,onerapidlygrowingstart-upthatChriswas involved with reserved a portion of each company meeting for oneemployeetogivea“gettoknowme”presentation.Thisallowedeveryonetogettoknowthe“newkidontheblock”onalevelwellbeyondthetypicalwelcomee-mail.Obviously,thisisthekindofactivitythatonlyworksattheTribestage
—inaFamilyitisn’tnecessary,andalargercompanywouldneverhavetimetointroduceeveryemployeeinthisfashion.When a company grows to the Village stage, simple logistics can make it
difficulttoholdacompanymeeting(oftencalledan“all-hands”meeting).Evenif the company hasn’t already grown to the point where it occupies multipleoffices,itmaybedifficulttofindaphysicalspacewherehundredsofemployeescangather.Rentinganoff-siteauditoriumforaweeklycompanymeetingisbothexpensive and impractical. The right thing is generally to shift the cadence ofsuch meetings to a lower frequency, such as monthly or quarterly, and toleveragetechnologylikevideoconferencingtobringthevariousofficestogether.One interesting approach is to have all employees use a teleconferencing
service rather than allow headquarters employees to have a better in-personexperiencethantherestofthecompany.Forexample,attheassetmanagementcompanyBlackRock,certainmeetingsareheldby teleconference,evenfor thesubset of employeeswho could gather in a single conference room so that allemployeesareonanequalfooting.Withtechnologyeliminatingthelogisticalchallenges,thecompanyall-hands
ishighlyscalable;thesebroadcasttechniquescanworkfororganizationsevenastheyscalethroughtheCityandNationstages.AtLinkedIn,youcouldtracethegrowth of the company over time by observing its all-handsmeetings.As thebusiness grew, these meetings moved from the company cafeteria to anauditorium,andtodaytheyinvolve livevideobroadcastsacross theglobe.Theall-handsneedstohaveaformalquestionsperiodsothatemployeescanrequestneeded information and feel part of the decision-making process.AtLinkedInwehavemoderatorsineachofficegatherquestionsformanagement.It’salsoatthispointthatthefounder/CEOneedstomakeaconsciouseffortto
develop broadcast channels to reach far-flung employees who might nototherwise feel a personal connectionwith the company’s leader.Of course, attheVillagestage,thecompanylikelyexceedsDunbar’snumber(thenumberofindividualswithwhom any one person canmaintain stable relationships), andthe founder simplywon’t have time tomeet one on onewith every employeewith any reasonable frequency. For example, even if you made time in yourschedule for two one-to-one employee meetings each day, and spent time ineveryoneofthecompany’soffices,youwouldstillonlytoucheachmemberofafive-hundred-personcompanyonceeveryeightmonths—notenoughtobuildastrongrelationship.
Switchingto“one-to-many”communicationsdoesn’talwaysfeelcomfortablefor founders andCEOs. PatrickCollison, the cofounder andCEO of the fast-growing payment company Stripe, described how he overcame these feelingswhenhevisitedourBlitzscalingclassatStanford:
The big change is the need for formal, explicit, broadcastcommunication. It feels unnatural, especially for me for somereason.Partofthewaytorationalizeitistorealizethatastart-upisnot a natural environment.The optimal things to do don’t alwaysfeelnatural.Thesocialgroupsyoubelong todon’t typicallygrow100 percent per year. The new people weren’t there for all thetortured discussions of the past. That can be good, but they alsodon’thavethecontext,soitisadelicatebalancingact.
BrianCheskyaddressesthisneedatAirbnbbysendingalonge-mailtoeveryemployee eachSunday night.Chesky’s e-mail isn’t simply a recitation of keyperformance indicators,whichcouldbe just aseasilyaccessedonadashboardsomewhere;rather,Cheskyshareshisthinkingonatopicheconsidersimportantto the company. This broadcast communication’s length, specificity, andauthenticitytransmittoeveryAirbnbemployeeanunderstandingofwhoCheskyisandwhatmatterstohim.Regular e-mails to all employees are a common best practice. Blitzscaling
masters Patrick Collison and YouTube’s ShishirMehrotra also employed thistechniquetomanagetheirrapidlygrowingorganizations.“Iwasabigbelieverinwritingaweeklyemail,”ShishirtoldourBlitzscalingclassatStanford.“Leaders[who]writethingsdowntendtodealwith[fewer]communicationsissues.Youhavetoclarifyyourthoughtprocessesinacompletelydifferentway.Ifyoujusthave a meeting and say, ‘Okay, so we’ve all decided,’ then people playtelephone.”If you can’t overcome your discomfort with writing, you can record and
distribute regular voicemails or short videos. These broadcasts can besupplementedwithsmallerone-to-manyevents,suchasQ&AsessionswhentheCEO visits a local office, or a breakfast with that month’s newly hiredemployees.Forexample,MarkPincusofZyngaholdsMondaymorningcoffeetalkswithallnewemployeesjoiningthatweek.Electroniccommunicationsaregreat for establishing regular contact, but face-to-face interaction is still
important for establishing a deeper, more emotionally resonant relationship.ReedHastingsmeetsthisneedbynothavinganofficeatall,andwanderingthehallsandconferenceroomsofNetflix.As your company grows and plays an increasingly important role in your
industry,youwilllikelyexperienceaneedtomakemoreofyourorganization’ssensitiveinformationsecret.Youprobablywon’tsharethebankbalancewithallemployees,orkeeppeopleuptospeedonthetwistsandturnsofthelatestfund-raising effort. More secretive cultures might make this move in the Tribe orVillagestages,butasacompanygetscloser tobeingpublicly traded,even themostopencultureshavetomakemovesinthisdirection.
TRANSITION#5:INSPIRATIONTODATA
“Whatistheroleofdatainscalingyourcompany?”InaninterviewwithReid,Jeff Bezos of Amazon discussed how he makes data a critical part of hismanagementprocess.“If this isadecisionbasedonopinions, thenmyopinionwins,”saidJeff.“However,databeatsopinion.Sobringdata.”Jefffollowsthispolicy faithfully; on one occasion, he argued that Amazon customers wouldneveranswerquestionsfrompotentialcustomersaboutaproduct.Justtoomuchfriction,he thought.Theproduct teamdidn’t try tochange Jeff’sopinionwithrhetoric and argument; instead, they emailed product questions to a thousandAmazon customers who had recently purchased a product and tracked theresponses.ThedatatheirsimpleexperimentproducedchangedJeff’smind,andthe“CustomerQuestions&Answers”sectionthatresultedhasaddedbillionsofdollarsinincrementalsalesbyincreasingconversionrates.Dataisthelifebloodofdecisionmakingforanycompany,butitisparticularly
fundamentalifitinformsthedesignofyourproduct,orifacquisitionmarketingis your key distribution strategy. For example, when he was at Twitter, myGreylockcolleagueJoshElmanneededtofigureouthowtokeepTwitterusersactivelyusingtheservice.Byanalyzingthedata,hewasabletodeterminethatthe“coreusers”whowere90percentlikelytobeactivemonthaftermonthwereusing Twitter on at least seven different days per month. Further analysisshowed thatwhatset theseusersapart fromthe lessactiveuserswas that theyfollowedover thirtyotherTwitterusers.OnceElmanunderstood thesefigures,Twitterwasable toencouragenewusers to followmoreaccounts,and,withinsixtydays,Twitterwasabletogetitsratioofdailyactiveuserstomonthlyactive
usersaboveits50percenttarget.Mostcompaniesstartwithrelatively little in thewayofanalyticsduring the
Family and Tribe stages (they might have performed an analysis to estimatemarketsize,buttheyrarelyhavemuchdatafromactualcustomers).Atthisstageyou’reintroducinganewproduct,notfine-tuninganexistingprocess.Youdon’tneed an analytics dashboard to know if people are using your product or not.Andifcustomersaren’tusingyourproduct,adashboardisn’tgoingtotellyouhowtochangecourse.Inotherwords,ifyoudon’thavecustomerstolistento,thebestyoucandoislistentoyourgut.ButasHarvardBusinessSchool’sRanjayGulatiandAliciaDeSantolanoted
in“Start-UpsThatLast,”thisapproachdoesn’tscale:“Improvisationisintegraltoyoungventures;it’showtheymakediscoveries.However,asfirmsgrowtheyneed a framework of plans and goals to guide them. Thatway they can keeptrying new things and reacting to dynamic markets, but with an eye towardlarger objectives and sustaining the business. Otherwise improvisationessentiallyamountstoaimlessriffing.”You’re already dealingwith your fair share of unknowns as your company
growsatabreakneckpace,soitmakessensetoseekcertaintywhereveryoucan.Tomakethetransitionfrominspiration(orimprovisation)todatamoreeasily,ithelpstostartwiththebasics.Trackafewkeystats,suchasthenumberofusers(registered users, application downloads, retail buyers, etc.), churn, and rawengagement.WhenSelinaTobaccowalajoinedSurveyMonkeyin2009,shehadtobuildupthecompany’sdatainfrastructurequickly.“Therewerenoanalyticsbefore2009,”SelinatoldourBlitzscalingclassatStanford.“Therewasadailycashreportandthatwasit.Istronglybelievethatasawholecompany,youcan’tgetbehindmorethanthreetofivemetrics.Thekeymetricswepickedwerefreeusers, free users that become paid users, and then user engagementmetrics—numberofsurveys,andreturnrate.”Sometimes even a single metric can tell you a lot. At YouTube, Shishir
Mehrotradecidedthattheirsingleclarifyingmetricwouldbewatchtime.“Ourgoalwastogettoonebillionhoursperdayofwatchtime,”hesaid.“Atthetime,we were doing 100 million hours per day. Facebook had about double that.Televisionasawholewas5.5billionhoursperday….Pickingasingleclarifyingmetric is very hard, but it clarifies decision-making and what constitutessuccess.”Whatevermetric(s) you select, that informationmust be easy to access and
provide clear context. Particularly when your company is still small and leanwith limited manpower, it pays to invest in the infrastructure necessary tosupportfast,data-drivendecisionmaking.Atext-basedlogfilemighttechnicallyprovideallthedatayouneed,butanyonewhohastomanuallyprocessthatdatainto a user-friendly graph each timewill rapidly stop using that data to drivedecisions.Whatmattersisn’twhatyoucollectbutwhatyouconveytodecisionmakers.The key statswill evolve as your companygrows.You can’t simply “set it
andforgetit”whenitcomestodata.Thecriticalmetricsforpredictingthelong-term viability of your business may be very different as you achieve scale,particularly if the environment is changing rapidly. For that matter, yourdefinition of “long-term” will change a great deal. In the Family stage, nextmonthoftencountsas“long-term,”whereasaNation-stagecompanymighthavemultiyearplans.AtLinkedIn,webeganwithalaserfocusonthenumberofuserregistrations as our key stat, but the long-term engagement of our users and anumberofotherstatsaremoreimportanttoday.Thisdoesn’tmeanthatyoushouldthrowoutallyouroldmetrics;therecanbe
a lot of value to continuity. For example,MariamNaficy ofMinted toldme,“Thekeyistocreateconsistentquestionsfromthebeginningandtonotchangethem over time, because that’s the only way to compare metrics over time.We’vebeenusingNetPromoterScore[acustomer-loyaltymetricthatmeasureshow likely customerswould be to recommend a product or service to others]fromthebeginning.”WatchoutforwhatEricRiesdubbed“vanitymetrics”—numbersthatpresent
arosypictureofthebusinessbutdon’tactuallyreflectitskeydriversofgrowth.Note that one company’s vanity metric might be another’s key driver. Forexample,pageviewsareavanitymetricformoststart-ups,butthekeydriverfora media company. In an interview for Reid’sMasters of Scale podcast, EvWilliams, founderofBlogger,Twitter, andMedium, reported that in the earlydaysofTwitter,histeamgotcaughtupinaparticularlyharmfulvanitymetric.Twitterwasbeingpraisedinthepressforencouragingdeveloperstobuildontopof itsAPI, andEv’s teamcelebrated the rapid rise in thevolumeofAPI callsTwitter was handling each day. Unfortunately, they discovered that API callvolumedidn’tactuallycorrelatewithbusinesssuccess.Infact,theoppositewastrue;thelargenumberofAPIcallswereoverwhelmingTwitter’sinfrastructureandcausingscalabilityandperformanceissues.“Wediscoveredthatalotofthedevelopers who built on top of our API were very inefficient,” he recalled.
“TherewasoneMexicanradiostationthathadaparticularlybadJavaScriptontheirWebpage—justthatoneWebpagewasbringingusdown!”TwitterhadtotightenitsAPIaccessrulestoreducethecallvolume.Whatevermetrics you choose,when the organization is still small, the data
cangenerally spreadviaosmosisbetween individualemployees, supplementedby a regular review during weekly companymeetings. You don’t need fancybusinessintelligence(BI)toolsoradedicatedteam.Once your organization reaches the Village stage, however, osmosis won’t
work.Yourpeopleareworkingonmultiplethreads,andtheorganization(whichhas exceeded Dunbar’s number) is now too big for everyone to know oneanother. Using a common dashboard will allow you not only to see how thethreads interlock but also to coordinate thework of different groups.Throughthedashboardeachgroupcan tell theothers, “This iswhatwe’reworkingon;thisishowwe’redoingit;andthisishowwe’reworkingtogetherwiththerestofyou.”AlmostallqualityVillage-sizebusinesseswilluseadashboard toassess the
daily health of their companies. Your organization’s dashboard will tell youwhatyouwanttotrackandensurethatyou’rekeenlyawareofsuddenchangesso that you can quickly investigate any surprises and assign actions to theresponsiblepersonorgroup.At the City and Nation stages, you’ll almost certainly need a dedicated BI
team to ensure that the necessary data is getting to the people who need tosupportandcarryoutkeydecisions.Thestakesaresohigh,andthecostofbaddecisionssogreat,thattheexpenseofadedicatedteamissmallincomparison.Mark Pincus invested heavily in his BI team at Zynga, which allowed the
companytotrackeveryclickintheirgamesratherthanrelyonGoogleAnalyticslikemostofhiscompetitors.“Peoplewouldsay,Zyngahasfiftypeopleworkingon analytics, this other company only has ten,” Mark recalled during aninterviewforReid’sMastersofScalepodcast.“Zyngamustbedumb.Actually,collectingthatdataletusmakeandevaluateourbetsfaster.”In addition to simply supplying data and insights to existing business units,
manyof the top-performing companies create adedicatedgrowth team,whichcombines marketing, product, and engineering to drive and coordinate theresponse to these insights. Most companies, even in the highly competitiveworldoftheconsumerInternet,stillthinkit’ssufficienttoconductalotofA/Btestsand iterateaccordingly.This isaneffective tacticbutpoorstrategy,since
local optimizations do not necessarily lead to a globally optimal result. Adedicated growth team can look at the big picture and see how product andmarketing decisions interact to produce (or not produce) the desired results.AccordingtoGreylock’sJoshElman,“Thebestgrowthteamsidentifythecoreinsightsthatgetusersfrom‘curious’to‘activatedhabitual’usersandbuildeveryfeatureandprogramintheproduct—includingthenonsoftwarefeaturesthatareapartofthewholeproduct—togetusersthroughthishurdlefaster.”Agrowthteamalsohelpsbymakinggrowthanumberonepriorityratherthan
asecond-orthird-classcitizen.Elmanlikestocompareatypicalmarketingteamto a Dickensian orphan, pleading with the product and engineering teams forresources:“Please,sir,mayIhaveanotherlandingpage?”Anyproductchangesorengineeringinfrastructureneededtodrivegrowth,nomatterhowpotentiallyvaluable,typicallyenduptakingabackseattotheproductorengineeringteam’sown road maps. In contrast, a growth team’s engineers can move far fasterbecausebuilding scalableandextensible testing infrastructure is a corepartoftheirjobs.Oneof thechallengesyoufaceasyoubuildupyourdatacapabilities is that
your strategy can disappear behind the numbers. The numbers might notmeasuretherealhealthofthebusinessorrevealtherealmajorthreatsyouface.Forexample,ifLinkedInweretoe-mailallofitsmemberseveryweektoremindthemtoupdatetheirLinkedInprofiles,theinitiativewouldresultinashort-termboost to profile edits. It would also be a horrible strategy, because it wouldannoytheusersanddegradetheuserexperience.JonathanRosenbergofGooglehastoldthestoryofhowblindlymanagingto
the numbers led Excite@Home astray. Excite@Home measured the click-throughs on every element of its home page. If an element didn’t look like itwouldhititsclick-throughtarget,Excite@Homewouldmaketheelementmorevisuallyprominent. Inotherwords, in attempting tohit its numbers, thehomepage teamwasemphasizing the leastcompellingelementsandde-emphasizingthemostcompellingones!Thisiswhyyoumayneedtoblendquantitativeandqualitativeanalysis.Our
friend John Lilly likes to distinguish between “genius-driven design” (e.g.,Apple) and “data-driven design” (e.g., Google). Both approaches have theirstrengths and weaknesses. Data-driven design is great at optimizing productswithincrementalchanges,butitcouldsteeryoutothetopofalocalhillratherthan the highest peak. Genius-driven design may be the only way to build a
revolutionaryproduct,butitusuallyneedstobesupplementedwithdata-drivenrefinement.
TRANSITION#6:SINGLEFOCUSTOMULTITHREADING
As the company grows, the product focus will also undergo a major change,fromasingle-threadingtoamultithreadingapproach.Whatwemeanbythisisthat start-ups in the early stages of blitzscaling are generally single-productcompanies that focus on doing one thing extremely well. But to keep thecompanygrowinginthelaterstages,scale-upsneedtomanagemultipleproductlinesorevenbusinessunits.We don’t know of a single start-up that succeeded without starting out as
single-threaded.Thatfocusisthekeytobeatinglargercompetitorsintheearlystagesofacompany’sexistence.Foryears,DrewHoustonofDropboxwastoldthatGooglewouldkillhiscompanybecauseof itssecretive“ProjectPlatypus”(which eventually launched as Google Drive). Houston found theseproclamationsmore annoying than frightening, because he knew the power ofsingular focus. In an interview for Reid’s Masters of Scale podcast, heexplained:
ForacompanylikeGooglethat’sdoingahundreddifferentthings,there’saverylongbreadlinetogetthenextgoodengineer.Andifyou’reproject#35,whichisaboutwhereGoogleDrivewasontheirlist,it’sgoingtotakealongtimebeforethatteamgetsfedwithanyamazingpeople.Whenyouconsidertheelevenplayersyouputonthefieldversusyourcounterpartatabigcompany,youcanactuallyhaveamassivetalentadvantage.NotbecauseGoogledoesn’thavegreatengineers;theyprobablyhavebetterengineersthanyou.Butthe leaderof theproject is amidlevelproductmanager forwhomit’s just the next rung on the ladder.As a founder, you’re just somuchmorecommitted,andyourteamissomuchmorecommitted.
Today, years after the launch ofGoogleDrive,Dropbox continues to grow intermsofbothusersandpayingcustomers—somuchforthe“Dropboxkiller.”Evencompaniesthatpivotseveraltimes,asPayPaldidinitsfirstyear,needto
stayfocused,especiallyastheyshifttheirattentionandeffortfromoneinitiative
toanother.MyGreylockcolleagueJosephAnsanelli,thecofounderandCEOofthecustomer service software start-upGladly, tells entrepreneurs, “Don’t tryasecond channel until you have your main flywheel working. Most successfulcompaniesdominateonechannel.”The shift to multithreading usually occurs during the City stage of
blitzscaling. Once the company has more than a thousand employees, theorganization is large enough to support the creation of multiple divisions orbusinessunits.Whilemovingtoadecentralizedorganizationmakesithardertocoordinate the different divisions or business units, the keymotivation for thechangeishowitallowseachgrouptofocusonitsspecific thread.Your teamsneedtheability—andthemanpower—torelentlesslypursueaspecificobjective;askinga team tosplit its timebetween twodifferentbusiness lines is likely toresultinthefailureofboth.This is especially true when the main thread is a business line that has
matured. In their Harvard Business Review article “The AmbidextrousOrganization,” Charles A. O’Reilly III and Michael L. Tushman draw thedistinctionbetween“exploiting”and“exploring.”Maturebusinesslinesfocusonincremental innovations that help them exploit awell-knownmarket,whereasnew threads focus on more radical innovations and exploring a new marketopportunity.Theyexaminedthirty-fiveattempts tospinupnewthreads,acrossninedifferentindustries.Whattheyfoundwasthattheseeffortsweremostlikelyto be successful in “ambidextrous” organizations,where the new threadswereorganized as structurally independent units but integrated into the existingmanagement structure. Inotherwords, the leadersof thenew threadsnotonlyhave the freedom to innovate but also the ability to coordinate with seniorleadership to leverage existing resources and expertise from more maturethreads.Multithreadingyourorganizationallowsyoutotackleproblemsthatmightnot
bevulnerabletoasingle-threadedapproach.AtLinkedIn,forexample,weknewthatweneededtoaddresstheissueofuserengagement.LinkedInisenormouslyvaluable as a databaseof résumés, but it is evenmorevaluable as the leadingcommunityforprofessionals.Thechallengewasfiguringouthowtodevelopadaily use case that helped LinkedIn users with their professional lives andencouragedthemtousetheservicecontinuouslyratherthanjustwhentheywerelookingtoswitchjobsorhireanewemployee.Wetriedanumberofsingle-threadedeffortstomeetthechallenge.Werolled
outfeaturesoneafteranother,suchasarecommendationengineforpeoplethatourusers shouldmeet andaprofessionalQ&Aservice.Noneof themworkedwellenoughtosolvetheproblem.Weconcludedthattheproblemmightrequirea Swiss Army knife approach with multiple use cases for multiple groups ofusers.Afterall,somepeoplemightwantanewsfeed,somemightwanttotracktheir career progress, and some might be keen on continuing education.Fortunately, LinkedIn had grown to the point where the organization couldsupportmultiplethreads.Wereorganizedtheproductteamsothateachdirectorof product could focus on a different approach to address engagement. Eventhoughnoneofthoseeffortsaloneprovedasilverbullet,theoverallcombinationofthemsignificantlyimproveduserengagement.Multithreadingcomeswithadefinitecost.Somepeopleareeagertojumpto
multithreading as quickly as possible because they think it increases theircompetitive bandwidth. In reality, you should be thoughtful and careful aboutmaking thisdecision.Companies likeGooglegrant agreatdealof freedom toindividual units, and, as a result, the different products and services donot fittogetherseamlessly.ManyofGoogle’sservicesarestrongenoughtosucceedontheirown,butthismeansthattheyaresucceedinginspiteof,ratherthanbecauseof,multithreading.In contrast,Apple’s highly centralized approach allows it to producehighly
integrated and polished products, but, as a result, it restricts itself to a muchsmallerproductline.Ofcourse,thisisintentional;SteveJobsalwayswantedtorunasclosetosingle-threadedaspossibletomaintainApple’sunityofpurpose.OneofthefirstthingsStevedidwhenhereturnedtoAppleasCEOin1997wasto reduce the company’s product line from dozens to a simple two-by-twomatrix: consumer desktop, pro desktop, consumer laptop, and pro laptop.“Decidingwhat not to do is as important as decidingwhat to do,” he told hisbiographer Walter Isaacson. Another famous Steve story involves an Applestrategy off-site where Apple’s top one hundred people worked for a day toreduceApple’sstrategytotenkeypriorities,atwhichpointStevecrossedoffthebottomsevenitemsandsaid,“Wecanonlydothree.”Generally, you should start adding threadswhen it’s strategicallynecessary,
andwith a realistic assessment of the negative impact thatmultithreadingwillhaveonorganizationalfocus,resourceefficiency,andsoon.AtLinkedIn,wemadeanexplicitstrategydecisiontomultithreadourrevenue
model, even though theconventionalwisdom inSiliconValley is to stick toa
single revenuemodel.Wewerecriticized forhavinga“mishmash”of revenuestreams,suchasprosubscriptions, job listingfees,andenterprise licensingforour recruiterproduct. It’s true that therewasacost to thisstrategy in termsoffocus, but I believed thatwe didn’t have enough information to pick a singlerevenuestreamandhaveitbesufficienttobuildourintendedscaleofbusiness.Multithreading to support multiple revenue lines both mitigated strategic riskandhelpedusgettoscale.One important technique for making this decision is to consider both the
magnitude of the opportunity as well as its potential for gain. If you have abillion-dollaropportunity,itmakessensetoinvestmoreresourcesandekeouta5percentgain($50million)thantogrowanascentmillion-dollaropportunitybyafactorof1,000percent($10million).Thisiswhyit’sgenerallybettertohaveyourtenbestpeopleworkingonasingleimportantprojectratherthansplittingthem to attack two different opportunities. For example, AdWords is such anenormousrevenuedriverforGooglethateventinypercentageincreasesmakeahugedifferencetothebottomline.Conversely,whenthepotentialforgainassociatedwithyourcoreopportunity
declines,multithreadingisoftentheanswertoattackbettergrowthopportunities.The company eBay can be thought of as a collection ofmarkets.While eBaymight have started with collectibles, multithreading to expand into differentmarkets like cars andapparelwas essential to reaching its current scale.Morerecently, Tencent’s creation of WeChat is an example of aggressivemultithreading.Assuming that youmake the decision tomultithread your organization, the
optimalmanagementapproachistothinkofeachthreadasadifferentcompany.For each thread, you’ll need to identify a leadership team (“cofounders”) andcreate an incentive structure that allows it to operate with a great deal ofindependence and reap the benefits of success, without making your currentmanagers so envious that it tears the organization apart. This is alwayschallenging!Furthercomplicatingmatters,peoplewiththeentrepreneurialdriverequiredto
makemultithreading successful usually want to start their own companies, orapply their skills to thecompany’smain thread.One thing thatcankeep theseemployees motivated is making the various threads discrete projects—theequivalent of “apps” running on the main thread’s “platform.” This makes iteasytoanswerthequestion“Whyshouldn’tIjuststartmyowncompany?”by
pointingoutthebenefitsofbuildingontheplatform.Thisstructurealsomakesiteasiertomanagemultiplethreads,sincetheindividualthreadsarelesslikelytocomeintoconflict.The incentives ofmultithreading have to reflect the success of each thread,
whilestillkeepingtheleadershipofeachthreadinvestedinthesuccessofalltherest. Without this balance, the different threads might engage in internecinewarfareoverresources,andtheindividualleadershipteamsmightprioritizethesuccessofasecondarythreadoverthehealthoftheentirecompany.Youwanttogiveleadershipareasontomakeeachthreadwork,butnotattheexpenseoftheothers; in otherwords, youwant the “owners” of eachof the threads thinkinglike anownerof theoverall company.Poorlydesigned incentives canmake itnearlyimpossibletoshutdownathread,evenifitsperformanceispoor,sinceitsleadershipmightfighttoothandnailtostayopen.Youmight be tempted to simply treat each thread like a separate company
within an overall holding company. After all, doesn’t that work for WarrenBuffett at Berkshire Hathaway? The difference is that Berkshire Hathaway’scompaniesareseparate,noncompetitive,cash-generatingbusinessesthathaveahistoryofindependentoperationsandcompletemanagementteams.Incontrast,when a blitzscaling company starts setting up multiple threads, they are stillattached,mightbecompetitive,arelikelyconsumingthesamepoolofcash,andhavenohistoryofindependentoperations.Oneof thepeople I’vepersonally seenhandle these issueswith exceptional
skill isDeepNishar,LinkedIn’s formerheadofproductandnowatSoftBank.Deep set up LinkedIn’s different product threads and expertly managed theproductleaderstocreateabroadersenseofownershipviaawebofalignment.Eachproduct leaderwas theownerof aprimary thread,butwasalsopartiallyaccountableandcompensatedforhisorherworkinsupportingafellowproductleader as a secondary thread. This produced an additional layer of alignment,which reinforced the alignment of all being part of the LinkedIn “holdingcompany.”
TRANSITION#7:PIRATETONAVY
This key transition is the shift from playing offense to playing offense anddefenseat the same time.Morepoetically, it’s the shift frombeingapirate tobeingpartofthenavy.Itrequiresanevolutioninstrategyaswellasanevolution
incompanyculture.For decades, technology entrepreneurs have had an affinity for pirates. As
withmanyoftheclassictropesofthestart-upworld,thelinkbetweenstart-upsand pirates was codified by the late Steve Jobs. Andy Hertzfeld, a legendaryserial entrepreneur who worked at Apple and helped design the originalMacintosh, related the story on hiswebsite Folklore.org.When Jobs gatheredtogethertheMacintoshteamforanoff-siteshortlyafterthereleaseoftheLisa,he famously kicked off the proceedings by laying out three “Sayings fromChairmanJobs”asguidingprinciplesfortheproject.
1. Realartistsship.2. It’sbettertobeapiratethantojointhenavy.3. Macinabookby1986.
InspiredbySteve’swords,theMacintoshteamcreatedahomemadepirateflag,completewith the classic skull and crossbones,with a rainbow-coloredApplelogo decal as an eye patch. The image of a pirate continued to be so widelyassociatedwithstart-upsthatwhenthecablenetworkTNTreleasedamoviein1999 about the heated rivalry between Steve Jobs/Apple and BillGates/Microsoft,itwastitledPiratesofSiliconValley.Therealityisthatmanystart-upsarelikepirates:theylackformalprocesses
andarewillingtoquestionandevenbreakrules.Thisflexibilityiscriticalintheearly stages of building a great company. Pirates don’t convene a committeemeeting to decide what to do when an enemy ship is approaching—they actquicklyanddecisively,andarewillingtotakerisksbecausetheyknowthatthedefaultoutcomeisdeath.Early-stage start-ups are alsoon the full offensive,wagingguerrillawarfare
on bigger, established competitors. They are used to striking quickly, usingsurprise as aweapon, and taking on risks that established companies can’t orwon’t.Duringtheearlystagesofblitzscaling—FamilyandTribe—it’seasiertotakerisksbecauseyoudon’thavemuchtolose.AsKrisKristoffersonwroteandJanisJoplin(amongothers)sang,“Freedom’sjustanotherwordfornothinglefttolose.”But if you succeed as a pirate, you’ll eventually win enough wealth and
territorytoblitzscaletotheVillage,City,andNationstages.Atthatpoint,eventhemostinveteratepirateswillhavetotradeintheirJollyRogerfortheflagofa
legitimate,disciplinednavy.If theydon’t, theirorganizationswilldevolve intochaos.EventuallyCaptainJackSparrowhastogrowupandstartactingmorelikethe
soberandresponsibleCaptainPicard.Thistransitioncanbechallenging.Foundersandearlyemployeesoftenresist
changingtheirapproach;afterall,didn’titbringabouttheirinitialsuccess?Plus,entrepreneurstendtohavearebelliousstreak;natural-bornrulefollowersdon’talways fare so well in a chaotic, “move fast and break things” start-upenvironment.Butfailingtomakethetransitionfrompirate tonavycanleadtodisaster.
ANoteonEthicalPiracy
Beforewegofurther,weneedtospendatleastalittletimeondispellingsomeof the connotations of the word “pirate.” In print and on-screen, pirates areportrayedinoneoftwoways:(1)lovableroguesand(2)sociopathiccriminals.The key differentiating characteristic of the lovable rogue, besides appearingmoreprominentlyonthemovieposter,isthatwhilesheorhemayquestionandbreak the lawsofpolite society,a lovable rogueadheres toapersonalcodeofethicsandtriesnottoharmothers.Thelovablerogueiswillingtobreaktherulesbut remainsmoral.He or she is an ethical or a “good” pirate. In contrast, thesociopathiccriminal,asthenamesuggests,behavesinapurelyselfishmanner,breaking rules and thoughtlessly harming others in order to bring materialbenefits.While start-ups and their founders may benefit from behaving like ethical
pirates,theyshouldneverbehavelikesociopathiccriminals.Besidesthefactthatsuchanapproachismorallywrong,asapracticalmatter,yousimplycan’tbuildaworld-changingcompanyasanoutlaw,anditisdifficulttomaketheshiftfromdeviant to mainstream society. And this is particularly true in a world wheresocialmediaisquicktoshineaspotlightonunethicalpracticesthatcantarnishacompany’s reputation forever. Go afoul of the law, and your customers willneitherforgivenorforget.One of the key ways to assess whether you’re being an ethical pirate or a
sociopathistoask,“AmItryingtochangetherulesforeveryone,orjusttryingtogetawaywithapersonalexemption?”AtPayPal,webroketherules,butwedidsobecausewewereworking towardabettersetof rules foreveryone.We
feltthatouractionswereethicalbecausewhilewecouldtechnicallyhavebeeninviolationof the letterofcertainbankingregulations(weconsistentlyarguedthatweweren’tabank,butnoteveryoneagreed!),webelievedthatinthelongrun we would be in compliance once we convinced the world to change therules, and that theworldwould be better off as a result.History demonstratesthatwewere right.Thevariousparties thatwereupset byour so-calledpiratementality—eBay, banks, regulators—all see the value of PayPal today. Bychanging the rules for everyone, we helped pave the way for other paymentcompanies like Square and Stripe, which have improved the world of mobilepaymentevenfurther.Rulesarenotholyscripture—theyexisttomaketheworldabetterplace,and
thusifyoucanimprovetherules,youshould.Ontheotherhand,rulesusuallyexist for a reason. You need to have some humility when breaking rules andrecognize that youmight not understand all the consequences. It’s not alwayscheatingtobreaktherules,butitisalwaysahigh-betaactivity,hencetheneedforcautionandcompassion.Apresent-dayexampleof a fieldwhere there areboth ethical andunethical
piratesistherapiddevelopmentofcryptocurrencieslikeBitcoinandinitialcoinofferings (ICOs) as a financing tool. The start-ups that are creating currenciesandholding ICOs are operating in a legal gray area and likely breaking rules.Someofthesestart-upsareethicalpirateswhoareworkingtochangetherulesforeveryone.Othersaresociopathiccriminalswhoaresimplytryingtocollectas much money as possible before the window closes and devil take thehindmost.Both typesmightmakemoney in theshort termif themarket ishotenough,butonlytheethicalpirateswillbeabletobuildlastingbusinesses,andonlytheethicalpirateswillhaveapositiveimpactontheworld.
JoiningtheNavy
WhenyourcompanyreachestheVillagestage,it’stimetostartthinkinglesslikeapirateandmorelikeanavy.What does thatmean?Well, you need to start following the rules, and you
might want to consider playing defense. Until now, your sole focus has beenoffense.Ifyoudon’thavecustomers,whydoyouneedtoworryaboutretainingthem? Now you should ask, “How can we lock out the competition?” Moreblitzscaling is often the answer. Being the first scaler helps you acquire
customers,lockininvestors,andattractthebesttalent.Iliketogeneratefresh,innovativewaystoplaydefensebyaskingmyteam,
“Ifwewere trying tocompetewithourselves,whatwewoulddo?What ifwewere a start-up? Google? Facebook? Microsoft?” You can also seek outsideperspectives, either by asking an independent boardmember or by leveragingnetworkintelligence.DuringtheCitystage,defenseoftenbecomestheprimaryfocus.Establishing
a new competitive edge tends to be very difficult. You should focus onstrengthening your existing market position instead. There are several bestpracticesfordoingso.First,trytoestablishastandard.OneoftheclassicSiliconValleyplaysisto
movefromanapptoaplatformsothatyoucanattractpeopletobuildonandtoyour platform (thereby leveraging the network effect of compatibility).Salesforce.com’sForce.com ecosystem is a great example of this.By offeringthe ability to build third-party applications on top of the Salesforce platform,Salesforcebenefitsfroma“forcemultiplier.”Thereareover2,800appsontheSalesforce AppExchange, and an International Data Corporation (IDC) studyshowed that the Salesforce ecosystem generates 2.8 times the revenues ofSalesforce.com itself. That means that while Salesforce.com has revenues of“only” $8.4 billion, its platformgives it the economic impact of a $32 billioncompany.Second,offeramorecompletesolution,andtrytooutflankthecompetition.I
liketosay,“Bothplayersareholdingglassesofwater,andaretryingtotipoverthe other person’s glass.” In other words, if your competitor suddenly startedoffering its core product for free, could you still make money on your coreproduct?Interestingly,thisfocusondefenseattheCitystageofblitzscalingisdifferent
inChinathaninSiliconValley.InChina,companieswillputteamsonanythingthathas traction; inSiliconValley, talent issopreciousand therearesomanyotheroffensiveplaysthatcompaniesoftencan’taffordtorelyonafast-followerstrategy.Thismeans that in a very real senseChina is evenmore competitivethanSiliconValley,thoughIexpectthatovertimeChinawillevolvetobecomemorelikeSiliconValleyinthisrespect.AttheNationstage,thetransformationfrompiratetonavyiscomplete.(Ifit’s
not,eitheryoudon’thaveaNationoryouhavefailedtomaketheshiftandyourNationisinchaos—witnessUberin2017.)
In this phase, acquisitions typically become important, if not essential, todefensivestrategy.Youcanacquireaninnovativetechnologyandteam,andthenfeedthemwithmassiveresourcesastheyscale.ThisishowGoogleblitzscaledAndroid. Google acquired Android in 2005, when it was still just a small,twenty-two-month-oldstart-upthatwasworkingonanewoperatingsystemformobile phones. Google let Android founder Andy Rubin hire additionalengineerstocompletetheproduct,whileusingitsmarketpowerandreputationtoestablish theOpenHandsetAlliance, a consortium topromoteAndroid thatincludedhardwaremakersSamsung,HTC,andMotorola,carriersSprintandT-Mobile,andchipmakersQualcommandTexas Instruments.With thisbacking,Androidgrewquicklyafteritslaunchinthefallof2008.AndroidexceededtheiPhoneinnumberofphonesshippedin2010,andatoveronebillionphonesperyear,todaymakesupnearly80percentofglobalsmartphoneunitshipments.Acquisitions are the biggest offensive and defensive plays in your Nation
playbook. Think about how certain key acquisitions won a major market fortheiracquirers.TheYouTube,Instagram,andWhatsAppacquisitionswerebothdefensiveandoffensive.AcquiringYouTubeallowedGoogletorecoverfromitsfailed Google Video initiative, but it also kept YouTube out of the hands ofcompetitors likeMicrosoft. The Instagram andWhatsApp acquisitions helpedFacebook defend againstmobile incursions, but they alsomade Facebook theleaderinmobile.Financial strategy can also become competitive strategy. For example,
Apple’scashhoardallowsittomovequicklyandpaycashforanyacquisition—twokeyadvantagesduringacompetitivebiddingprocess.Finally,youmayorderyournaval taskforces to launchdiversionaryattacks
thatyieldlittletacticaladvantagebutthathelptheoverallstrategicsituation.Forexample,MicrosoftneedstofieldasearchenginetocompetewithGoogle,eventhough it isunlikely tocapturemuchmarket share,becauseGoogle is fieldingproductivityappsagainstMicrosoft.Atthisphase,youshouldtrytomakeyouropponentsdefendeverybitoftheirterritories,because,ifyousucceed,theywillbestretchedtoothintowardofftheattacksyouactuallyconsiderimportant.Just remember to save a few ships to fend off attacks from those pesky
pirates!
FromCaptaintoAdmiral
At the time of the writing of this book, the ridesharing company Uber wasSilicon Valley’s most valuable start-up (and second globally to its frenemy,China’s Didi Chuxing), despite having spent most of 2017 in the news for anumberofseriousproblemsandscandals.Someoftheseissueswereduetoclearlyunethicalbehavior,includinginternal
problems,suchas thesexualharassment reportedby the formerUberengineerSusan Fowler, and various external attempts to subvert free competition,regulation,andthepress,suchascreatingfakeaccountstopoachdriversfromitsrivalLyft(asreportedbyTheVerge),developingsoftware(Greyball)topreventlawenforcementandregulatorsfromaccessingtheservice,andthen-COOEmilMichael suggesting that the company spend money to hire oppositionresearcherstointimidatejournalists.This kindof behavior is unacceptable, regardless of the size or stageof the
companyundertakingit,andhasrightfullybeenwidelycondemned.YetevenifUberhadneverengagedintheunethicalbehaviorsoutlinedabove,
the company would still have faced real issues because of its reluctance toabandon its pirate-like strategies (many of them benign in its earlier days)despiteitsmuchgreatersizeandscope.WhenUber’sboardpickedDaraKhosrowshahiasthecompany’snewCEOin
September2017,itcertainlyhelpedthatDarahadawell-deservedreputationforrunningano-dramaoperation(aclassicnavalofficer, inotherwords).But justas important was his experience in successfully growing Expedia into aprofitable $20 billion twenty-thousand-employee giant that has won praise asoneofthebest-managedcompaniesinitsindustryandagreatplaceforwork-lifebalance.WhileDarawillbedealingwithmanyflashy,well-publicizedissuesatUber,
hisbiggestchallenge—andgreatestopportunity—willbesteeringUberthroughthedifficultbutcriticaltransitionfrom“pirate”to“navy.”Tocreateafriendliercultureandhaltthemassdefectionoftoptalent,winbacktheloyaltyofdriversand riders, andput an end to the legalbattles that haveplagued the company,Uber’snewchiefexecutivewillneedtostartbehavingmorelikeanadmiralandless like a pirate captain.All start-ups recognize that there is a value to beingsmall: innovation, nimbleness, focus, outcome versus process. All successfulentrepreneurshavethedesiretostaysmallinthisway.Butthemostsuccessfulscale-ups are those that have managed to keep the positives of staying smallwhilereapingthebenefitsofbeingbig.
DarawastryingtostrikethisbalancewhenheresetUber’sculturalnormsinNovember2017.HeannouncedthechangesinapostonLinkedIn.
Aswemovefromaneraofgrowthatallcoststooneofresponsiblegrowth, our culture needs to evolve. Rather than ditchingeverything, I’m focused on preserving what works while quicklychangingwhatdoesn’t.This is the approachwe’ve takenwithournewcultural values,
whichwe announced to employees today.Our values definewhowe are and howwework, but I had heard frommany employeesthatsomeofthemsimplydidn’trepresentthekindofcompanywewanttobe.
Afirmbelieverthatculturemustbewrittenfromthebottomup,Daradidn’tcomeupwithanewsetofvaluesalone,behindthecloseddoorofaconferenceroom. Instead,heaskedemployees tosubmit ideas forhowto improveUber’sculture. Over 1,200 people sent in submissions that were voted onmore thantwenty-twotimes.ThenewculturalnormsthatDaraunveiledreflecthisdifferentapproacheven
inthesimplenatureoftheirlanguage.Insteadofhighlighting“lonewolf”mottoslike“Alwaysbehustlin’,”thenewcultureemphasizesthegroupbystartingeachnormwiththeword“We”:
Wecelebratedifferences.Wedotherightthing.Weactlikeowners.Wemakebigboldbets.
Dara deserves credit for working hard to add “navy” values such asresponsibility and doing the right thing to Uber’s “ethical pirate” values ofboldnessandaggression.But cultural change,while necessary, is not sufficient to turn a pirate gang
into a real navy.When the CEO of a large organization likeUbermakes thetransitionfromcaptainofasinglepirateshiptoanadmiralrunningafleetwithnaval-likediscipline, there arewell-established techniques andapproaches that
can help make this transition smoother and more effective. For example, ifyou’rebuildingaglobalbusiness,therearethreekeyelementsyouneedtoputinplace.
1. A set of managers who are responsible for, and have strong executivecontrolover,theirindividualmarketsglobally
2. Anunderstandingofhowthosemarketsdiffer,which leads toavarietyofplansforhowtogrowineachofthosemarkets
3. A unified executive team to coordinate global operations, including theactivityoftheindividualmanagersleadingoperationsineachcountry
Thefirsttwoelementsinvolveadecentralizedcommandstructurethatallowstheindividual“captains”oftheshipsinthefleettooperatewithentrepreneurialvigor. The third involves a centralized staff that can help the “admiral”coordinatetheactionsofthefleetformaximumimpact.Uber actually did a good job with the first two elements. Uber’s general
managersarelikeindividualshipcaptains,andtheirabilitytoactindependentlyhelpedUberdevelopinnovationslikesurgepricing(whichwasanindependentexperimentconductedintheBostonmarket).WhereUberfailedwasitsinabilitytocommittothethirdelement,aunifiedexecutiveteam.Whenyouhavestrongindividualcaptainsandanadmiralwhocan’torwon’tbuildastafftohelphimorheractuallymanagethefleet,youendupwithapiratemob.The failure to build a unified executive team is sadly common. Some
entrepreneurs find it difficult to accept the increased structure and decreasedfreedom of a formal staff; many of these people started companies preciselybecausetheydislikedthefeelingofworkinginalargeorganization.InhisbookonUber,WildRide,thejournalistAdamLashinskydescribeshowUber’sTravisKalanickviewedhisroleatthehelmofhisgiantcompany:
“ThewayIdoit,itdoesn’tfeelbig,”[Kalanick]says,fallingbackon a favorite trope: that he approaches his day as a series ofproblemstobesolved….“Iwouldsayyouconstantlywanttomakeyour company feel small,” he says. “You need to createmechanisms and cultural values so that you feel as small as
possible.That’showyoustayinnovativeandfast.Buthowyoudothat at different sizes is different. Like when you’re super small,you go fast by just tribal knowledge. But if you did tribalknowledgewhen you’re super big it would be chaotic and you’dactually go really slow. So you have to constantly find that linebetweenorderandchaos.”
Kalanick’s words reveal a pirate’s discomfort with running a largeorganization.“HeobviouslythinksofhimselfastroubleshooterinchiefasmuchasaCEO,”Lashinskywrites.Butwhileactingasatroubleshooterinchiefmightbeagoodfitforhispersonality,attheCityorNationstage,gettingtooinvolvedinthedetailsofindividualproblemsisprobablyapooruseofaCEO’stime.Kalanick, inotherwords,wasdoingwhat feltgood tohimrather thanwhat
theorganizationneeded.The purpose of hiring a management team is to solve the organization’s
problemsinamorescalableway.TheCEOshouldbethehub,andtheexecutiveteamthespokesthatconnecttheCEOtothefrontlinemanagersandemployeesoperatingwheretherubberhitstheroad.Kalanickwastryingtobethehubandthe spokes rather than helping the organization build the ability to get thingsdonewithouthispersonaloversight.Anothersymptomof thisdysfunctionwasKalanick’s habit of canceling his executive staff meetings. Without spendingtimetogether,itisdifficultforamanagementteamtobuildagroupcultureortocoordinate the many initiatives of the organization. A strong executive teammeetsonaregularbasisandfocusesonthemostimportantinitiativesandissues,including active planning for the future. According to a 2018 Forbes article,“Inside Uber’s Effort to Fix Its Culture Through a Harvard-Inspired‘University’,”Uber’sSVPofLeadershipandStrategy,FrancesFrei,describedthemanagement team’s lackofcohesionasoneof the largestproblemsfacingthecompany.Thearticlereportsthat“Uber’sseniorexecutivesweren’tworkingas a team and only had one-on-one relationships with Kalanickwho oversawthemall.”Kalanick is absolutely correct when he argues that staying small helps
organizations stay innovative and fast, but staying small isn’t always apossibility. It is better to build an organizational structure that you can iteratemultiple timesrather than toavoidscaling theorganizationas longaspossibleandmaketheshiftinasinglegiantleap“someday.”
In other words, you have to build management strategies that scale. EvensomeoneassmartasLarryPagelearnedthisduringtheearlydaysofGoogle;hetriedtorunGoogle’sengineeringdepartmentwithoutmanagementbyhavingallfour hundred employees report directly to then-VP of engineering WayneRosing.The failureof this experiment convincedhim to allow then-CEOEricSchmidttobuildarealorganizationalstructureatGoogle.Anygivenmanagement structure is likely tobe temporary.Youcan’t run a
Village thesamewayyourunaTribe,andyoucan’t runaCity thesamewayyourunaVillage.Butwithoutstructure,youwon’tmakeittothenextstageofgrowth.It appears that Kalanick’s discomfort with Uber feeling “big” led to a
dysfunctionalorganizationalstructureinwhichheclungtohispreviousways.Intheabsenceofacohesivemanagementteam,UberseemedtooperateonamodelthatSusanFowlerdescribedinherpersonalblogas“agame-of-thronespoliticalwar”withmanagersfightingforadvancement:
Theramificationsofthesepoliticalgamesweresignificant:projectswereabandonedleftandright,OKRswerechangedmultipletimeseachquarter,nobodyknewwhatourorganizationalprioritieswouldbeoneday to thenext,andvery littleevergotdone.Weall livedunder fear that our teams would be dissolved, there would beanother re-org, andwe’dhave to start onyet another newprojectwith an impossible deadline. It was an organization in complete,unrelentingchaos.
WhenUber tried to scale itsmanagement by hiring experienced executiveslike Jeff Jones fromTarget, they ended up resigning rather than changing theorganization. During the first half of 2017 alone, Uber lost eight VPs ordepartmentheads.In contrast, companies like Facebook and Amazon, and leaders like Mark
ZuckerbergandJeffBezos, foundways tosuccessfully recruit leadership fromtheoutside,blendingthemwithexistingteammemberstochangeandstrengthentheorganization.Facebookpromoted insiders likeChiefProductOfficerChrisCox(whojoinedFacebookasasoftwareengineerin2005afterdroppingoutofStanford), but also brought in compatible outsiders like Sheryl Sandberg andMike Schroepfer. Jeff Bezos’s top lieutenants like Jeff Blackburn and Andy
Jassy areAmazon lifers, but he also brought in key outsiders like JeffWilkefrom AlliedSignal and former Chief Information Officer Rick Dalzell fromWalmart.Theseoutsidehirescanhelpevenatmassivescale;oneofthebenefitstoMicrosoftofbuyingLinkedInhasbeenaddingJeffWeinerandCTOKevinScotttoMicrosoft’sexecutiveteam.Asyour fleet of pirate ships and followers grows, youneed to intentionally
shapethemintoadisciplinednavy.Afleetofshipsrequiresstrongcaptainsanda strong centralized staff that can coordinate and harness their entrepreneurialvigor.Everysuccessfulfounderandeverysuccessfulorganizationmustgothrough
these changes. But as Uber has discovered, blitzscaling makes these changessimultaneously harder (because of the speed at which theymust happen) andmore important (because of the risk inherent in investing in speed overefficiency).
TRANSITION#8:SCALINGYOURSELF:FOUNDERTOLEADER
Allfoundersneedsomeuniversalskillstosucceed.Theyneedtheabilitytotakebold risks inpursuitofavision that isn’t self-evident toothers.Theyneed theabilitytolearn(sincethey’retryingtodosomethingbrand-new).Andtoplayalong-termroleattheirstart-upturnedscale-up,theyneedtheabilitytolivewithandresolvetheinevitableparadoxesofbeingafounder.WhenIaskedDropboxfounderDrewHoustontolookbackonhisexperience,hetoldme,“Ithinkalotofentrepreneursstartwithalotofinsecurityaboutwhattheydon’tknow.Whatyou want is not to be paralyzed by it, but to harness it—to use that nervousenergy to learn and make yourself better. You’ve got to keep your personallearningcurveaheadofthecompany’sgrowthcurve.”Maintaining a certain humility and a sense of perspective can help you
navigate thechanges inyour roleasyoublitzscaleyourcompany. Ifyou trulywant to blitzscale, then speed has to take priority over everything—includingyourownego.Thereareonlythreewaystoscaleyourself:delegation,amplification,andjust
plainmakingyourselfbetter.
Delegation
Canyoufind,hire,andmanagegoodpeople,thentransferworkovertothemsoyoucan tackle thechallengesyou’reuniquelysuited to tackle?Manyfoundersare so talented that they have a hard time letting go of tasks once they startperformingthem.Theyoftenthinkthingslike“WillsomeoneelsebeabletodothisaswellasIcan?”Theanswerisalmostcertainly“No,especiallynotatfirst,butthey’llprobablyfigureitoutovertime,justlikeyoudid.”Start-upsgetoff thegroundthanksto the individual talentandhardworkof
founderslikeMarkZuckerbergandBrianChesky,buttheyblitzscaleintogiantcompanies like Facebook and Airbnb because these founders learn how todelegate.One of the most important aspects of delegation, and often the most
challenging for a founder, is to hire an executive and hand off functionalleadership. For example, a lot of great founders are product people. Initialproduct/market fit and success are achieved because of their product instincts.But as the companygrows, these founderswill almost always need to hire anexecutivetotakeoverleadershipoftheproductorganization—it’stooimportanttobeafounder’spart-timejob.Akey technique Iuse toovercome this challenge is topicture thehire as a
specificliving,breathingpersonratherthanasarolewrittendownonapieceofpaper.Whenyoutrytopictureanabstract“headofproduct,”forexample,youmighthaveahardtimevisualizingthisfacelessentitydoingabetterjobthanyouare.Butwhenyoupictureaparticularindividual(say,JoeZadehofAirbnb),allofasuddenyourmindshifts to thinking,“Wow,just imaginehowawesomeitwould be to have someone like this running our product team.” It might bedifficulttohirethisparagon—executiveswhoarethatgoodarehardtopryloosefromtheircurrentcompanies—butitdoesn’thurttotry,andattheleast,you’llhave a great reference to which you can compare the people you actuallyconsiderhiring.
Amplification
Rather than delegate work you’re doing to others, can you hire people whoamplifytheworkyoudo?Thegoalhereisn’ttofreeyouupfromyourworksothat you can do other things; it’s to make the things you do much more
impactful.ThisisactuallyoneoftheareasI’vetriedtodevelopandrefineinmyownlife.Likemany founders and executives, I have an amazing executive assistant,
SaidaSapieva,tohelpmewithschedulingandlogistics.ButI’vediscoveredthatyoucantaketheconceptofamplificationmuchfurther.Forexample,Iwasoneof the first start-up leaders in Silicon Valley to borrow the “chief of staff”concept from the realm of politics and established corporations. Unlike atraditional assistant or even a technical assistant, your chief of staff shouldamplifyyourbusinessimpact:heorsheshouldbeabusinesspersonwhocannotonlymakecertaindecisionsforyoubutalsotriagetheimportantdecisionsthatyouhavetomakeyourself.Achiefofstaffcanalsomakesurethatallthepeoplewhowanttomeetorinteractwithyouare“briefed”inadvancesothatyourtimetogethercanbeasefficientandeffectiveaspossible.Myfirstchiefofstaff,BenCasnocha,was a successful author and entrepreneur beforewebeganworkingtogether;mysecond,DavidSanford,hadworkedwithmeatLinkedInandhadalsobeenanentrepreneur(andarestaurateur!).ItturnedoutthatBenandDavidwerebetteratorganizingmyownlifethanIwas;I’vebecomesignificantlymoreproductivesincetheystartedamplifyingmyefforts.Tolearnmoreabouttheroleandvalueofachiefofstaff,IrecommendthatyoureadBen’sessayonthetopic,“10,000HourswithReidHoffman,”whichyoucanfindonhispersonalwebsite,Casnocha.com.Onceyoubegintoappreciatethepowerofamplification,youcanfindmany
ways to scale yourself. For example, one of the things you need to do is toprocess information about your company, your industry, and the world as awhole.Ihaveafreelanceresearcheronmyteam,BrettBolkowy,whohelpsmelearnnew things and answerkeyquestionsby finding thebest informationonanyparticulartopic.Anotherkeyteammember,IanAlas,helpsmewithcreativeprojects like thevisualsummaries Iprepare formybooks.TheslideshowshecreatedformybookTheStart-upofYouhavebeenviewednearlyfifteenmilliontimes.Nowthat’samplification!NoramIuniqueinthis.Forexample,MarkZuckerberghasasubstantialteam
to help himmanage his socialmedia communications so thatwhen he travelsandmeetspeople,hecanmaximizetheimpactofhisinteractions.Trustedemployees,freelancers,orevenateamofoutsideconsultantscanbe
your amplifiers. The official nature of the relationship is less important thanhavingassistancethatyoucantrust.
MakingYourselfBetter
Because your company grows and changes so quickly as you blitzscale, it’scrucialforyoutofigureouthowtomakeyourselfbetterjustasquicklysothatyoudon’tbecome thebottleneck thatholdsyourcompanyback.Asour friendJerryChenlikestosay,“Therearenojobdescriptionsforfounders.Iftheroledoesn’tchange,there’ssomethingwrong.”Sinceyou’regoingtofacenewchallengesduringeverystageofblitzscaling,
youhavetomakeyourselfintoalearningmachine.MyfriendElonMuskisagreat example.He dropped out of Stanford’s PhDprogram in applied physicsbecause he thought he could learn more on his own! He started SpaceX andTeslabylearningliteralrocketscienceandcarmaking.Sohowdoyouaccelerateyour learningcurve so that youcan learnmore faster?Thekey is to stand, asIsaacNewtonwrote,“ontheshouldersofgiants.”Thismeanstalkingwithothersmartpeople,often,sothatyoucanlearnfrom
their successes and failures. It’s usually easier and less painful to learn fromanother’s mistakes than from your own. When I need to learn about a newsubject, I’ll definitely devour some books on the topic, but I almost alwayssupplement this reading by seeking out dialogue with leading experts in thefield. Brian Chesky at Airbnb, another amazing learning machine, doessomething similar, seeking advice from mentors like Sheryl Sandberg andWarrenBuffett.Brian toldour class atStanford, “Ifyou find the right source,youdon’t have to read everything. I’vehad to learn to seekout the experts. Iwantedtolearnaboutsafety,soIwenttoGeorgeTenet,theex-headoftheCIA.Even ifyoucan’tmeet thebest,youcan readabout thebest.”Brian lives thisadvice;hegotmanyofhisideasbyassiduouslyporingoverbiographiesofgreatentrepreneurslikeWaltDisney.Another helpful approach to seekingmentorship is to get help from experts
whomightbelessfamousthantheSherylSandbergsoftheworld,butwhohavefaced (and solved) similar issues in the recentpast. In an interview forReid’sMasters of Scale podcast,Dropbox’sDrewHoustondescribedhowhe tries tolearnfromfellowentrepreneurswhoareonthesamejourney:
Talkwith other entrepreneurs.Not just famous entrepreneurs, butpeoplewhoareoneyearahead,twoyearsahead,fiveyearsahead.You learnverydifferentand important things fromthosekindsofpeople. It really helps to have a sense of the longer-term arc,
becausethegamechangesquietlyfromphasetophase.
Inadditiontoseekinghelponanadhocbasis,Ibelieveit’sagoodideatobesystematicaboutlearningfromothers.Iadviseentrepreneurstohaveapersonalboardofadvisersor“boardofdirectors”whocanprofferadviceandhelpyoufill thegapsinyourknowledge.Forexample,Ihaveasetofinformaladviserswho helpme learn about the areas that matter tome, including very specifictopics like virality or people management. If you’re serious about somedayblitzscalingacompany,youshouldthinkofyourmentorsasaboardofdirectors.Regularlyreporttothemonyourprogress,andaskthemhowyoucandobetter.Everyone needs feedback. Brian Chesky, for example, likes to say, “I’mshameless about getting feedback.” He and I have a scheduled dinner everymonthwhere (among other things) we share what we’ve learned and providefeedback.Leveragingaboardlike thiscanhelpyoumanagerisksandincreasethepotentialupsideofyouractions.Thismaysound likea lotofwork,but it’s important to leaveyourself time
andspaceforreflectionandfeedback.It’seasytogetcaughtupinanendlessto-dolistandtolosesightofwhatisimportant.That’soneofthethingsIlearnedfromMarkZuckerbergandSherylSandberg.MarkandSherylmeet first thingeveryMondayandattheendofeveryFriday—nomatterhowbusytheyareorwhat else has come up.TheFridaymeeting is especially important because itgivesthemtimetolookbackovertheweekandreflectonwhatthey’velearned.Youmightfeellikeyoucan’taffordtotaketimeoutfromyourbusyschedule
tomakeyourselfbetter.Afterall,youmightthink,everyoneiscountingonme.This feeling, while natural, is counterproductive. Netflix CEO Reed Hastingswarned our Stanford class, “[When I was running Pure Software,] I felt likeinvesting inmewas selfish. I thought, ‘I shouldbeworking.’ Iwas invited tojoinYPO [YoungPresident’sOrganization], but I thought, ‘I can’t take a dayoff.’Iwastoobusychoppingwoodtosharpentheaxe.Ishouldhavespentmoretimewithotherentrepreneurs. I shouldhavedoneyogaormeditation. Ididn’tunderstand thatbymakingmyselfbetter, Iwashelping thecompany,even if Iwasawayfromwork.”Plus,whenyoumodelthebehavioroftakingthetimetoimprove yourself, you help encourage the rest of the company to develop acultureoflearning.
NINECOUNTERINTUITIVERULESOFBLITZSCALING
Blitzscalingacompany isn’t easy; if itwere, everyonewoulddo it.Likemostthingsofvalueinthisworld,blitzscalingiscontrarian.Tosucceed,you’llhavetoviolatemanyofthemanagement“rules”thataredesignedforefficiencyandriskminimization.Infact,toachieveyouraggressivegrowthgoalsinthefaceofuncertaintyandchange,youneedtofollowanewsetofrulesthatflyinthefaceof what is taught in business schools and are completely counterintuitive toaccepted “best practices” of either early-stage start-ups or classic corporatemanagement.
RULE#1:EMBRACECHAOS
Annual plans. Revenue guidance. Traditional business strives for order andregularityinmanagement,operations,andfinancialresults.Thisdesirefororderand regularity makes sense, because it allows companies to fine-tune theirapproachtobeasefficientaspossible,andgivesshareholdersapleasingsenseofstability.Butwhenyou’reblitzscaling,you’reexplicitlychoosingtosacrificeefficiency for speed, which means that the traditional focus on order andregularityneedstobereplacedwithauniquewillingnesstoembracealevelofchaosthatwouldhorrifymostHarvardMBAsandtheirprofessors.When you start a company, almost everything is an unknown, from the
product/market fit, to the competitive landscape, to the composition of yourfutureteam.Thereisnowaytoeliminatealloftheseuncertaintieswithcarefulplanning; most can only be resolved by doing. As a result, you have to takeactionevenifyouknowyoustillhaveissuestoresolve(andsometimesevenifyou don’t yet know exactly what those issues are). For example, manyentrepreneursstartbuildingproductbeforetheyhaveago-to-marketstrategy.Yet simply throwing up your hands is unlikely to bring success; passively
succumbing tochaos isnotawinningstrategy.Embracing chaos,on theotherhand, means accepting that uncertainty exists and therefore taking steps tomanage it. Ifyouknowthatyou’llmakemistakes, theanswer isn’t tositbackandwaitforanswerstofindyou,norisittochargeaheadwithoutpreparationorforethought.Youcan stillmake smartdecisionsbasedonyourestimateof theprobabilities, even without certainty. And, perhaps most important, you canmakesurethatyouhavetheabilitytocorrectyourmistakes.Myearlierbook,TheStart-upofYou,introducestheusefulconceptof“ABZ
planning.”EntrepreneursshouldalwayshaveaPlanA,aPlanB,andaPlanZ.
Plan A is your best current plan; Plan B is an alternate plan, based on the“adjacentpossible”towhichyoucanpivotifPlanAisn’tworkingoryoulearnofanevenbetteropportunity;PlanZisyourfallbackplanforsurvivingaworst-case scenario.ABZplanninggives youmultiple opportunities to recover frommistakesorsetbacks.Atmyfirststart-up,SocialNet,weweredelightedwhenwemanagedtohirea
brilliantserverengineer(PlanA).Thatdelightturnedtohorrorwhenheaskedtodeferhisstartdateforayear!Needlesstosay,astart-upcan’tsimplyputitselfonholdforayear—evenifyouhadthemoneytowaitoutthedelay,thelossofmomentumwouldprobablyconvincemostoftheteamtoquit.Wekeptlookingforotherbrilliantserverengineers(PlanB),butwhenwecouldn’tfindthem,wekept building the service anyway by asking othermembers of the team to dotheirbest,knowingthatwe’dhavetorebuildtheservicelater(PlanZ).Even if you do manage to hire the people you want, you’ll often have to
scramble their roles and job titles as the organization changes in response tomarketfeedback.AtPayPal,wethoughtwewereamobileencryptionproduct,and we hired accordingly. Then we shifted rapidly in turn to cash onmobilephones,thencashonPalmPilots,thenpaymentsbetweenPalmPilots,andfinallypaymentsviae-mail.Wecouldn’thavedonesoifourpeopleweretiedtoneatandtidyjobslike“mobileencryptionengineer.”TakeJamieTempleton,oneofourkeyearlyemployeesatPayPal.Wehired
Jamietoworkontheproduct,butoverthecourseofjustthreeyears,heshiftedfrom product to engineering to systems to policy, depending on what thecompanyneeded. Jamie is exactly thekindof employeeyouneed in the earlydays—someonewhoiswillingtoembracethechaosofastart-up—whichiswhyImadesurehejoinedmeintheearlydaysofLinkedInaswell.
RULE#2:HIREMS.RIGHTNOW,NOTMS.RIGHT
For most of Silicon Valley’s history, the conventional wisdom on hiringexecutivesintoastart-upwastoquicklybringinanexecutivewhocouldscale.Thismeanthiringsomeonewhohadexperiencewithmuchbiggerorganizations,theideabeingthattheirexperiencewouldcomeinhandyatalaterstage.In today’s start-up world, this rule no longer applies. The Darwinian
competitionissofiercethatyourorganizationneedstobe“allin”onthecurrentstageofscaling.Youneedmanagersandexecutiveswhoare“justright”forthe
currentphaseofgrowth;afterall,youwon’thavetoworryaboutthatnextphaseif your team can’t actually get you there. Hiring someone who has beenmanaging a thousand people to run a ten-person company is actuallycounterproductive,becausetheskillsneededtosucceedduringthosetwophasesareverydifferent.The ideal, of course, is to hire executives who can not only excel at your
currentphasebutstretch tocover thenextphaseaswell.But that“scalability”shouldbeasecondaryconcern.Theprimaryconcerniscurrentvalue.Youcanworry about whether to scale or replace an individual executive when thecompanyapproachesitsnextphase.Forexample,entrepreneursaresometimesadvisedtoavoidhiringsalespeople
untiltheyareabletosecureaVPofsaleswhohasshownthatheorshecanscalethe company to $100 million in sales. This is hogwash. The salespeople youneedtoignitehypergrowtharetotallydifferentfromthesalespeopleyou’llneedat scale.When you’re trying to sell your product for the first time, you needaggressive,adaptablesalespeoplewhoaren’tbigonfollowingrules.Bythetimeyou’veachievedscale,you’llneed thorough,process-orientedsalespeoplewhocankeepamachinerunningsmoothly.You’renotgoingtofindonepersonwhoisgreatatboth.Onethingto lookforwhenevaluatingapotentialhire iswhether theperson
seemsself-awareofwhichstagesoftheprocessheorsheexcelsatandprefers.For example, somepeoplewill tend togravitate toward early-stage companieswhere they will have more opportunities to take on a broad portfolio ofresponsibilities.Othersmight prefer early stages because they enjoy the directandtangibleimpactofbeingakeyindividualcontributororanimportant teamleaderover tackling theverydifferent andmoreabstractworkofbeinga full-timemanagerorexecutive.I’veknownanumberoftalentedpeoplewhopreferjoining early-stage companies because while they don’t want to take on thechallengeofbeingfounders,theydowanttobe,inthewordsofAaronBurrinHamilton,“intheroomwhereithappens.”Veryfewpeopleexcelatbeinganindividualcontributor,amanager,andan
executive, and even those rare employees are likely to have a preferred role.SeasonedSiliconValleyprofessionals tendtobeawareof theirpreferredstageandrolebecausethedisproportionateprevalenceofblitzscalingcompaniesgivespeople a chance to experience more of the different stages. These repeatedexperienceswithdifferent stages letemployeeszero inon thebest fit for their
skillsanddesires.Part ofhiringMs.RightNowalsomeansknowingwhen to let someonego
when themomentpasses.Forexample,agreatdesignermightexcel runningaone-womanshowataFamilyoraTribe,butbelesseffectiveworkingaspartofalargerdesignteam.AtLinkedIn,oneof thekeyemployeeswhofit thedescriptionofMs.Right
NowwasMinnaKing.Minna is an incredibly accomplishedprofessionalwhohascarvedoutavaluablenicheataveryspecificstageinthelifeofastart-up.Yousee,Minnaspecializesintakingasuccessfulsoftwareproductandhelpingitgoglobal.Shehasaveryparticularsetofskills thatshehasacquiredoveralongcareerdatingbacktothedot-comera.Sheknowsexactlywhatasoftwaredevelopment andproduct teamneeds todo inorder tomake Internet softwarework in different languages and markets, in areas ranging from databaseschemas to user interface. She then works with a cross-functional team toimplement these changes in advance of a global rollout. It’s not easy to findpeoplewhofityourneedssoperfectly;youcan’tjustgoonLinkedInandfilterby“preferredstageofblitzscaling.”(Thoughcometothinkofit,thatmightnotbe a bad idea…) You’ll probably have to rely on your network forrecommendations, which is where your investors and board of directors canhelp. But when you find Ms. Right Now, she can add huge value to theorganization.This is precisely what Minna did for me at LinkedIn, just as she did for
OvertureandeBaybeforejoiningLinkedIn,whichisexactlythesamethingshedid for two other highly successful software companies, SurveyMonkey andNextdoor,aftersheleftLinkedIn.Ineachcase,shecametothecompanyattheearly Village stage because in order for her work to add the most value, thecompanyneededtobebigandsuccessfulenoughtoneedtoglobalize,butsmallenoughtonothavetheinternalskillstodoso.
RULE3:TOLERATE“BAD”MANAGEMENT
When blitzscaling, speed is more important than having a “well-run”organization.Undernormalcircumstances,youshouldstrivefororganizationalcoherence and stability. Chaotic, unstable organizations make employeesnervous and hurtmorale. Butwhen you’re scaling up at lightning speed, youmayneedtoreorganizethecompanythreetimesinasingleyear,orrepeatedly
churn throughmembersofyourmanagement team.Whenyourorganization isgrowing300percentperyear,youmighthavetopromotepeoplebeforethey’rereadyandthenswapthemoutiftheysinkratherthanswim.Youdon’thavetimeto be patient and wait for things to “work out”; you have to act quickly anddecisively. There’s always a lot of change, and much of it isn’t voluntary.You’re building teams and the company simultaneously. In the interests ofspeed, youmight even surprise or blindsideyour people to cut down the timerequiredtomakeandimplementimportantdecisions.Problemsrelatingtojobtitlesareonecommonsymptomofthismessiness.In
the Family and Tribe stages, you don’t have time for a careful promotionprocess, and you don’t have time to sit around debating whether someone’sbusinesscardshouldread“headofengineering”or“seniorVPofproduct”(nor,forthatmatter,doyouhavetimetodesignandorderbusinesscards).Youmightjust keep employees’ titles the same even as they fail to reflect organizationalprogress and level of responsibility, or you might do things that no rationalcompanywoulddo,suchasdeliberatelyinflatingjobtitlestokeeppeoplehappyand counting on the ability to correct the situation “later.” Eitherway, you’retakingonorganizationalriskinexchangeforbeingabletofocuseffortswhollyongrowth.Consider the example of PayPal. While PayPal was a great success, the
companywas badlymanaged—and Iwrite that statement as one of its seniormanagers.Wedidafewgoodthings,suchasmakingsurethateveryemployeehadaclearprimaryjobandstayingfocusedwhenworkingoncertainimportantprojects,butforthemostpartPayPal’smanagementwasalackofmanagement.Therewere no one-on-one career development conversationswith employees.Therewasnoworkdonetoformteamsbeyondsimplypickingwhowasgoingtobelong to them. The few ruleswe hadweremore about individual incentivesratherthanteammanagement.Forexample,whenpeoplewerelatetoameeting,thelastpersontoarrivewasfined$100toenforcediscipline.Yetwhileweknewmeetingswereimportant,wedidn’tdesignateanotetakertocapturekeypointsandactionitems,acommonandbasicpracticeinSiliconValley.But PayPal’s “bad” management provided a number of counterintuitive
strengthswhilewewereblitzscaling.DuringthecriticaltimeswhenPayPalwasdeveloping its businessmodel innovations and scalingup,we foundourselvesneeding to navigate a series ofmake-or-break challenges, or, as I like to callthem,“Ohshit!”moments.
Oh shit, we have a fraud problem and we’re losing millions of dollars wedon’thave.Ohshit,Visasayswehavetochangetheproductorthey’llshutusdown.Oh shit, eBay,ourmost importantbusinesspartner, just started itsownventuretodirectlycompetewithus.Becauseofour“bad”management,wedidn’thaveanypreconceivednotions
of“thisiswhatthecompanymustlooklikeinthreeyears.”Thechaoticnatureof our management actually kept us nimble in the face of these serious,unexpected landmines.When everyone in the organization has roles that areundefined and in flux, it’s easier to say, “I know this is what you’ve beenworking on for the past four days, but nowwe’re doing something different.”Theinternalchaoshadtheeffectofnormalizingradicalchangeforourpeople,whichmeant theywerebetter able to adjust to the radical changes theoutsideworldwasthrowingatus.Weknewthatwewereslalomingthroughaminefieldwhileotherpeopleshotatus.ToparaphraseBruceBanner/theIncredibleHulkfrom themovieThe Avengers, the secret behind our superpowerwas that wewerealwayschanging.Wewerealsofortunateinourtiming.Onethingthatholdsteamstogetherin
the absence of management is an opportunity to win. After the dot-com bustbegan, a lot of tech companies were failing, but PayPal still had a chance tosucceed.Allyouhadtodowastolookatthechartshowingthecontinuingrisein daily transactionvolume!Soour people put upwithmore than theywouldnormallyputupwithbecausetheywantedtowinandlikedbeingpartofateamofhigh-powered,high-IQplayers.Classic “good” management and planning presume a certain amount of
stability that isn’t always available when you’re blitzscaling. One of themisconceptionsofentrepreneurshipisthatyouworkoutaplanandthenexecuteit.Thinkoftheembeddedmetaphorin“building”abusiness—theverylanguagesuggests that you’re following an architectural plan.Butwhenyou’re creatingand scaling an innovative business model, you often don’t have any detailedblueprints.Instead,it’smorelike“Ithinkabuildingovertherewouldbeagoodidea.Let’sstartdigging!”Thenoncethecementispouredandthewallsgoupyourealize,“Itshouldbeahotel,andthereforeweneedtodothiskindofafloorplan.”Is that “bad”management?Maybe.But if badmanagement saves you from
buildingawarehouse in thewrongpartof townand letsyouquickly turn thatstructureintoasuccessfulhotel(orsavesyoufromlosingmoneyonmobilecash
and letsyouquicklycapture themarket forglobalpayments), then itmightbethebestapproachyoucantake.
RULE4:LAUNCHAPRODUCTTHATEMBARRASSESYOU
It’snot thatyoushouldstrive toproduceabadproduct.Rather, ifyouneedtochoosebetweengettingtomarketquicklywithanimperfectproductorgettingtomarket slowly with a “perfect” product, choose the imperfect product nearlyevery time.Getting tomarket fastallowsyou tostartgetting thefeedbackyouneed to improve it. Any product that you’ve carefully refined based on yourinstincts rather than realuser reactionsanddata is likely tomiss themarkandwill require significant iteration anyway. The ideal is a tight OODA loop—observe, orient, decide, act—over and over again. Speed really matters, andlaunchingearlyletsyouclimbthelearningcurvetoagreatproductfaster.MarkZuckerbergcreditsspeedfor thesuccessofFacebook. Inan interview
forReid’sMastersofScalepodcast,Marktoldus,“Learnandgoasquicklyasyoucan.Evenifnoteverysinglereleaseisperfect,Ithinkyou’regoingtoendupdoingbetteroverayearortwothanyouwouldbeifyoujustwaitedayeartogetfeedbackonallofyourideas.Thatfocusonlearningquicklyisthefocusofthecompany.”I learned this lesson the hard way when I was running my first start-up,
SocialNet.Ididn’twanttobeembarrassedbyourfirstrelease,sotheapproachwe tookwas to complete the entire product beforewepulled back the curtainandletpeoplesignup.ThisapproachdelayedSocialNet’slaunchbyayear,andwhenwe finallydid launch,wequickly realized thathalfof the featureswe’dpainstakingly implementedweren’t important, andhalf of the important thingsthat our service would be useless without were missing because we hadn’tthought of them. While there were other reasons why SocialNet failed, notlaunchingearlyanditeratingbasedonmarketfeedbackwasprobably themaincauseofdeath.After my experiences at PayPal, and the success we found through rapid
launchesandproductiteration,IwasdeterminedtolaunchLinkedInassoonaspossible.Ourteamdefinedalistoffeaturesthatwethoughtweretheminimumrequiredtoenterthemarket.Yearslater,SteveBlankandEricRieswoulddubthis a “minimum viable product” (MVP). For LinkedIn, theMVP included auser’sprofessionalprofile,theabilitytoconnecttootherusers,asearchfunction
tofindotherusers,andamechanismforsendingmessagestofriends.Shortlybeforelaunch,westartedworryingaboutwhetherLinkedInwouldbe
usefulwithoutacriticalmassofprofiles. Ifauser logged in toLinkedIn,howcouldwemake itusefuleven ifnoneof thatuser’s friendshadsignedupyet?WedecidedthatwhatwasmissingwasaContactFinder,aversionofsearchthatwouldletaLinkedInuserfindpotentialvendors.Forexample,ifyouneededaconsultant tohelpyoufigureouthowtoglobalizeyourservice,youcoulduseContact Finder to find Minna King. Our engineering team estimated that itwouldtakeusamonthtobuildthisfeature.Wewerepresentedwithadifficultchoice—delay the launch by a month, or launch without a feature that wethought might be essential to our success. Operating on the embarrassmentprinciple,welaunchedwithoutContactFinder.Andquicklywediscoveredafarbiggerproblem:UnlikeusersofpersonalsocialnetworkslikeFriendster,whichwere growing explosively as new users invited their friends to join, LinkedInusers weren’t sending any invites. Our user growth was stalled. Our baselineproductwasembarrassingbecausenoonewasusing it! Ifwehaddelayed thelaunchamonth tobuildContactFinder, therestillwouldn’thavebeenenoughpeople hanging around to use it, meaning that we would have lost a monthbuilding a feature that didn’t address the core problem.We estimated thatwewouldneedatleastonemillionusersbeforesearch(andContactFinder)wouldbeuseful,andsolvingthatproblemwasthetoppriority.Basedonthelaunchdata,wefocusedontryingtoincreasevirality,whichis
howwe became the first social network to allow you to upload your addressbook.ThisfeaturehelpedLinkedIngettoacriticalmassofoveronemillionuserprofiles,andtherestishistory.Keep inmind that you should be embarrassed by your initial release—not
ashamed or indicted! The desire for speed is not an excuse to cut dangerouscorners.Ifyoutriggerlawsuitsorburnthroughyourmoneywithoutlearning,itmeansyoudidlaunchtoosoon.Thepointoflaunchingyourproductearlyistolearnasquicklyaspossible.Butyour learning isuseless ifyoudon’thave theability to iterate. If your product bursts into flames and kills someone, youprobablywon’tgetanotherchance.ThefirstlaunchofLinkedInfellwellshortofourexpectations,butwedidn’tdoanyharm.Beforeyoureleaseyourproduct,make sureyouknowwhatyou’re trying to learn, andhowmuch riskyou cantakewithoutendangeringyourcustomersoryourreputation.Entrepreneurshavetowalkafinelinebetweenfixableandfatalflaws!
The line between fixable and fatal often depends on the nature of yourproduct. If we consider just two dimensions of product—free (or freemium)versus paid and consumer versus enterprise—each combination can be placedalongacontinuum:
A free consumer product can get away with the most flaws, becauseconsumerstendtobeverytolerantwhenitcomestosomethingthatdoesn’tcostthemanything.
A free enterprise product needs to bemore refined; even if it is free, thestakesarehigherinaprofessionalsetting.
A paid enterprise product needs to be evenmore refined, but it can stillhave significant flaws, because these types of products are intended forexpertuserswhomayhavenochoicebuttousetheproduct.
Apaidconsumerproducthastheleastroomforerror.Whileconsumersareverytolerantofflawsinfreeproducts,theyexpectproductstheypayfortobe nearly perfect andwill grouse loudly about any significant flaws theyfind:“WhatamIpayingforhere?”
Sometimes you can reduce the risks and uncertainties by obtaining userfeedback without actually launching. Design thinking often calls for rapidprototypingandusertestingthroughpaperprototypesorvisualizationtoolslikeInVision and testing tools like UserTesting.com. Yet even these techniquesadheretotherule—thegoal is totestasearlyaspossibleratherthantrytogetthingsrightbeforeunveilingthemtousers.Onceyou’velaunchedyourproduct,youhavetomakesureyoulearntheright
lessons from market feedback. As the example of LinkedIn’s initial launchshows, the key lessonsmight not be found inwhat your customers say but inwhat theydo.The firstusersofLinkedInwere largelyour friendsand family,andtheydidn’ttellus,“Thiscrapisuselesswithoutmoreusers!”Instead,theytoldusthingslike“Itseemslikeitwillbeprettyuseful”—yeteventheyweren’tsendingoutatonofinvites.Yes,youneedtolistencarefullytowhatyourusershavetosay,butyoualsoneedtoknowwhentoselectivelyignorethem.Whenanecdotaluserfeedbackanddatacontradicteachother,listentothedata.Peopleareoftenquitebadatpredictinghowthey’llreacttochanges.Thescientifictermis the inconsistency between predicted and observed behavior. For example,
when Facebook was considering adding a feature that would use facialrecognitiontoautomaticallytagmembers’facesinphotographs,thefocusgroupparticipantswere very negative toward the concept, calling it “creepy” and aninvasionofprivacy.YetwhenFacebooktestedthefeature,auto-taggingboostedengagementanduserslovedit!WhenIofferthisadvice,Isometimesheartheobjection“That’snottheway
SteveJobsdid it.”Well,holdonaminute.Firstofall,contrary to thepopularnarrative, not all of Steve’s productswere perfect from the start. The originalMac didn’t comewith a hard drive. The original iPhone didn’t comewith anApp Store. It is true thatwe can point to a number of entrepreneurswho didlaunch a great product at the very beginning. For example, when ElonMusklaunchedtheTeslaModelS,itimmediatelybecamethehighest-ratedcarontheroad,beingnamedMotorTrendCaroftheYearinitsdebutyear,andachievingahigherConsumerReportsratingthananyothercarthatorganizationhadevertested.Buttodothis,youhavetobelievethatyoucannailtheproduct/marketfitof a newmarket before you launch, and invest substantial amounts of capitalbased solely on that confidence. Elon bet his own fortune, and hundreds ofmillionsofdollarsofinvestorandgovernmentmoney,thatTeslacouldbuildabettercarthananyofhiscentury-oldcompetitors.Thenumberofentrepreneurswhoareableandwillingtobetsoaggressivelyislow.Thenumberwhocandososuccessfullyisevenlower,andthenumberwhocandoitsuccessfullymorethanonceislowerstill.So,yes, ifyouarea raregeniusandcanaccuratelyandconsistentlypredict
whatthemarketwants,trustingyourinstinctswillbefasterthanusingtrialanderrortoiterateyourwaytoabetterproduct.Goodluckwiththatapproach!Asameremortal,Iprefermarketfeedback.
RULE5:LETFIRESBURN
Ioftentellentrepreneursthatstartingacompanyislikejumpingoffacliffandassemblinganairplaneonthewaydown.Thedefaultoutcomeforanystart-upisdeath,whichmeansthatyouhavetomovequicklyanddecisivelytoavoidthatdefaultoutcomeatallcosts.Thatdoesn’tleavealotoftimefordottingeachiorcrossingeacht.Ateverystageofblitzscaling,therearealwaysfarmoreproblemsandissues
clamoringforyourattentionthanyouhavetheresourcestoaddress.Youmight
feel like a firefighter, except instead of trying to extinguish a blaze in onecontainedspot,youcanseeseparate firesallaroundyou—andyoudon’thavetimetoputoutallofthem.Oneofthewaysthatblitzscalingentrepreneurscanstayalive isbydeciding to let certain firesburn so that theycan focuson thefires that if allowed to rage unchecked really will destroy the company. MyGreylock colleague Joseph Ansanelli says, “What you say ‘no’ to is moreimportantthanwhatyousay‘yes’to.”You can’t ignore those fires forever—they are actually dangerous and will
eventually require attention, but they aren’t relevant at most points duringblitzscalingbecauseextinguishingthemdoesn’tmovetheneedleontheexpectedoutcome.Pictureanemergencyroomsurgeontryingtosavethelifeofatraumapatient;asshe’sconductingemergencysurgery,shemightnoticeasuspicious-lookingmass,butshe’sgoingtofocusonpatchingthepatient’sarteriesfirst—therewillbetimeforbiopsiesandtestslater.Afterall,ifthepatientdiesontheoperatingtable,evenapotentialtumorwillbeirrelevant.Theart,ofcourse, isknowingwhichfires to letburn.Prioritizingyourfires
tendstobeafunctionofacombinationofdifferentfactors.Thefirstisurgency:Which fire is going todamageor kill yourbusiness the soonest?This doesn’thavetobelimitedtofiresthatendangertheexistenceofthebusiness;forastart-up,afirethatkillsyourabilitytogrowisnearlyasdeadlyinthelongrunasonethatthreatenstoputyououtofbusinesstomorrow.Usually,yourfirststepistodecidewhether you can simply punt on the problemand tackle it later.WhenSelinaTobaccowalajoinedSurveyMonkey,oneofthefirstfiressheconsideredfighting was the design of the product. It was ugly, outdated, and franklysomewhat embarrassing. But it was also extremely effective and successful—userengagementwasgood,andcustomerswerehappy.Selinadecidedtodeferredesigning the product in favor of more urgent fires. This decision made itharder for her to recruit aesthetically sensitive engineers, but it didn’t kill thecompany.Insomecases,ifthefiresatyourstart-upareburningmoneybutnottouching
the customer, and if you’re able to afford the waste, you might be able toliterallybuy time and ignore them.Frequently, raisingmore capital is an easy(thoughoftenexpensive)waytokeeplessurgentfirescontained.Thesecondfactoryouwanttolookatisefficacy:Whichfiresdoyouhavethe
abilitytoextinguishrightnow,andwhichwillbeeasiertoextinguishlater(andviceversa)?Ifafire isurgent,butyoucan’teffectivelyfight it rightnow,you
mighthavetoignoreitandhopethatexternalcircumstancesputitout.Likewise,if it isn’t necessarily urgent right now, but will wreak a lot more havoc ifallowedtospread,youmightconsidersavingyourselftheordeallaterbynippingitinthebud.Thefinalfactortoconsiderisdependency:WillextinguishingFireAmakeit
easier to extinguish Fires B and C? These knock-on effects can be veryimportant, because there are alwaysmore fires burning than you have time tofightatanygiventime.I believe that there is a Maslovian hierarchy of fires that applies to most
rapidlygrowingstart-ups,wherethetopofthelististhemostimportantfiretofightfirst:
DistributionProductRevenuemodelOperationsCompetitionWhat’snext?
What this means is that for most consumer Internet start-ups, the mostimportant fire is distribution; if your distribution goes up in flames, yourcompanyisdoomed.Ifyouareable tocontain thatfire,however, itwillmakefightingtheotherfiresawholeloteasier.Acquiringusersgivesyoufeedbackonhow to improve your product. Acquiring millions of users or thousands ofcustomersmakesitaloteasiertogeneraterevenue.Generatingrevenuemakesiteasier to pay for the infrastructure and personnel to scale up your operations,eitheroutofcash floworby raising investment.And ifyouhavea successfulandgrowingbusiness,thenitmakessensetoworryaboutthecompetition.In the case of LinkedIn, after we had fixed our distribution problem by
buildinginviralityandgeneratingasignificantuserbase,wehadpeopleharpingontherevenuemodelfire.IfIreceivedanickelforeverytimesomeoneaskedme, “How is LinkedIn going to makemoney?” during those days I probablywouldn’thaveneededanotherrevenuemodel!ButIknewthatweshouldignorethatfire,because(1)thelackofrevenuewasn’tgoingtobetheproximatecauseofdeathunlessitpreventedusfromraisingmoneyand(2)theproductfirewas
far more urgent and required our focused attention. If we couldn’t find thedistribution to acquire a critical mass of at least a million users, and build aproducttheyfoundcompellingenoughtobecomeregularusersoftheservice(oratleastrespondtoLinkedInrequests),therevenuemodelwouldbeirrelevant.Atthetime,potentialSeriesAinvestorswantedtoseeabusinessmodelthat
showedhowLinkedInwouldget toprofitability. I toldpotential investors thatweweren’tgoingtogeneraterevenueuntilafterthenextroundoffunding,andthatthereforeitshouldn’tmattertothem.Theyinsistedanyway,sotheteamandI generated a financial model that included revenue sources. I don’t evenrememberwhatweputinit!Ratherthanwasteweeksonit,wesimplysetasideasingleevening,drankacoupleofglassesofwine,andputtogetherthemodel(Imighthavebeenalittlemiffedathavingtospendevenasingleevening,butitwasprettygoodwine,soitwasn’tatotalwaste).This story also highlights why you need people on your team who have a
tolerance forchaos, risk,anduncertainty.Mostofusarewilling to fight fires;it’s a smaller subset of people who are capable of noting the presence of aroaring blaze that might soon cut off all escape routes without allowing it todistract them from their laser-focused effort to fight an evenmoreurgent fire.ThemembersoftheLinkedInteamwerecomfortablewiththatuncertainty,andcould still work at full effectiveness even though we didn’t have a definedrevenuemodel.Plus,ifyourpeoplecan’tletfiresburn,they’llspendalloftheirtime fighting them, which won’t leave any time for coming up withbreakthroughopportunitiestoadvancethebusiness.
RULE#6:DOTHINGSTHATDON’TSCALE(THROWAWAYWORK)
PaulGraham,thecofounderofYCombinator,wroteafamousessayinwhichheadvised entrepreneurs to do things that don’t scale. This advice is spot-on foryoungstart-ups,butit’sevenmoreimportantforblitzscalingstart-ups.Engineershatedoingthrowawaywork.Notonlyisitwasteful,itoffendstheir
senseofefficiency.Theyarefirmbelieversintheconventionalwisdomthatsaysit’sbettertobuildyourproductrightthefirsttime,soyouonlyhavetobuilditonce.Butwhenyou’reblitzscaling,inefficiencyistherule,nottheexception.Toprioritizespeed,youmightinvestlessinsecurity,writecodethatisn’tscalable,andwait for things to startbreakingbeforeyoubuildQA toolsandprocesses.It’struethatallofthesedecisionswillleadtoproblemslateron,butyoumight
nothavealateronifyoutaketoolongtobuildtheproduct.Ahackthattakesatenthofthetimemaybemoreusefulthananelegantlyengineeredsolution,evenifithastobethrownawaylater.Muchthesamelogicapplies tonearlyeveryaspectofyourbusiness.You’ll
often have to do things that don’t scalewhen it comes to sales (e.g., founderMarcBenioffbroughtinSalesforce.com’sfirstcustomer,BlueMartiniSoftware,by calling in a favor from its CEO Monte Zweben), operations (e.g., PaulEnglish listedhis personal cell phonenumber as the original customer servicelineforKayak),andsoon.Noristheworldneatlydividedinto“thingsthatdon’tscale”and“thingsthat
scale,”with the former smoothly—andpermanently—givingway to the latter.Thecodeorprocessthatscalesduringonestageofblitzscalingmaybreakdownattheverynextstage,andwhateveryoureplaceitwithmightnotscaleatfirsteither.ConsiderhowthefoundersofAirbnbsolvedtheproblemofhostspostingpoor-quality photos of their rental properties onAirbnb.com: they became thephotographers.AsBrianCheskytoldme,“WewouldborrowcamerasfromourRISD[RhodeIslandSchoolofDesign]friendsinBrooklyn,thenliterallyknockonthedoorsofallourhosts.”Together,BrianandcofounderJoeGebbiacouldphotographabouttenhomes
perday(cofounderNathanBlecharczykhadtostayattheapartmentthatdoubledas theiroffice,makingsure thesitedidn’tcrash).Talkaboutdoing things thatdon’tscale!Once,ahostaskedBrianwhenhe’dgetpaid,andBrianpulledthecompanycheckbookoutofhisbackpackandwrotehimacheck.“Iguessyou’renotaverybigcompany,”thehostsaidashepocketedthecheck.AsAirbnbtookoff,thephotographyfunctionhadtoscaleupconsiderably.So
thefoundershiredphotographersfromCraigslist,hituptheirRISDfriends,andeven recruited Airbnb hosts who listed photography as a hobby. By tappingthese sources, the company was able to build a stable of five to tenphotographerswhowerepaid$50perhome,andwhomthey trackedusing thesophisticated management tool of a spreadsheet with photographers and theirassignments.Prettysoon,thissystemtoowasoverwhelmed.SotheyhiredEllieThieleasa
summerinternfromSyracuseUniversity,andmademanagingphotographersherfull-timejob.Byfocusingsolelyonmanagingthephotography,Elliewasabletoincrease the number of active photographers to about fifty. Itwas only at thispointthatAirbnbwenttoatrulyscalablesolution:software.Nathanwrotesome
code,addingtwobuttonstothesite;oneforhoststorequestaphotographerandthe other for Ellie to trigger a payment when a photographer finished anassignment.EventuallythefoundershiredJoeZadehasanentry-levelengineerandaskedhimtoworkwithEllietofullyautomatethephotographyprocess.Airbnbworkeditswaythroughthreedifferentwaysofhandlingphotography
before building any code, and has rewritten the photography systemmultipletimessincethen.Itwouldn’thavemadesenseforAirbnbtostartbybuildingascalableautomatedphotographysystem;at thepointwhen thecompanybeganthis journey, the site was receiving a mere ten visitors per day, and the onlyengineeringresourcewasNathanBlecharczyk.AnyworkhedidonthisproblemwouldhavedelayedalltheotherengineeringworkAirbnbneededtogetdonetogrow its business.By doing things that didn’t scale, the companywas able togrow despite the resource constraints and the “wasted” work of buildingspreadsheetsthatwouldhavetobethrownawaylater.
RULE#7:IGNOREYOURCUSTOMERS
The fundamental rule of customer service has long been “The customer isalways right.” But for many blitzscaling companies, the key rule is “Providewhatever customer serviceyou can as long as it doesn’t slowyoudown…andthatmaymeannoservice!”Manyblitzscalingstart-upswilloffere-mailsupportonly, or no support at all, relying on users to find and help one another ondiscussionforums.Onanabsolutescale,ignoringyourcustomersisrarelygoingtobeapositive.
Customers like to feel heard, and ignoring them will eventually deplete yourcompany’ssupplyofgoodwill.Butforblitzscalingcompanies,lettingcustomersfeel ignored isoftenoneof the fires that’seasier foryou to letburnuntilyouhavefinishedfightingthebigger,moredeadlyfires.Our experiences at PayPal offer a telling example of how hypergrowth
requiresrapidchangesinyourapproachtocustomerservice.InFebruary2000,transaction volumewas increasing 3 to 5 percent per day, on a compoundingbasis.Eachdaywewerefallingbehindtothetuneofthousandsofunanswerede-mails, which compounded the problem because the users who didn’t get aresponsetotheirinitiale-mailwouldsimplywriteinagain.Conventionalwisdomwouldhavecalledforus todevoteasmanypeopleas
possible tocustomersupport.But that’s theoppositeofwhatwedid.Outofa
forty-person team, we had two support people (and our office manager wasspendinghalfofhistimetohelpout).Wehadmuchmoreurgentfirestofight.Forexample,duringthatsametimeperiod,wewere(1)raisingourfirstmajorround of venture capital, (2) starting to compete with Billpoint, our biggestpartnereBay’sattempttocloneourbusiness,and(3)negotiatingamergerwithElonMusk’sX.com.Sufficeittosaythatthingswerebusy,andwedidn’thavethe bandwidth to solve the customer service problem. So we ignored ourcustomers!Afterall,noneoftheircomplaintsstoppedtransactionvolumefromgrowingexponentially.Ofcourse,ignoringourcustomershaditsowncost.EventhoughPayPalwas
onlylistedinthelocalPaloAltophonedirectory,enoughpeoplelookedupthenumber and dialed random extensions that at any time of day, every phonewouldberingingwithanangrycustomerontheotherend.Westoppedpickingupthephones.Ignoring customers is a temporary solution. Eventually, after we raised a
significant roundofventure capital andhadannounced theX.commerger,wehadthetimeandresourcestodealwiththeproblem.Weforgedanalliancewiththe governor of Nebraska and announced that we were hiring for customersupportpositionsinOmaha.WhyOmaha?X.comalreadyhadasmallcustomerservice team located there.Why had X.com picked Omaha? One of its earlyemployees had a sister living there who offered to help the fledgling start-uptakecustomerservicecalls.Weendedupflyingoutmostofthecompanytoholdgroupinterviewssowe
couldhireandtrainonehundrednewsupportemployeeswithinthirtydays.ThePayPalemployeewehiredtoleadthecharge,SarahImbach,endedupmovingtoOmahaforeighteenmonths.Fortunately,therewasahappyendingtothestoryforeveryoneinvolved:Ourproductwasusefulenoughthatourcustomersstuckwithusuntilwewereabletostartservingthem.WebeatBillpoint,wentpublic,and ended up selling the company to eBay for $1.5 billion.As for Sarah, hereighteen months in Omaha were productive on several levels; in addition tobuildingaserviceandoperationsorganizationthatstillemploysoverathousandpeopleinOmaha,shealsometherhusbandthere.
RULE8:RAISETOOMUCHMONEY
Entrepreneursgenerallytrytoavoidraisingmorecapitalthantheyneed.Raising
excessiveamountsofcapitaldilutestheirstakeinthecompanyandintroducesapreferenceoverhang(allthatmoneyhastobepaidbacktoinvestorsbeforethefoundersandemployeesgettoparticipateintheupside).Yetwhenblitzscaling,youshouldalwaysraisemore—preferablymuchmore—thanyouneed.“Excess” cash allows you to better account for the unforeseeable—and the
only thing that’s foreseeable about blitzscaling is that you will at some pointencounter theunforeseeable.That includesanythingfromastockmarketcrashoroutlandishexpenses to anopportunityyoucouldn’tpredict in amarket thatdidn’texistwhenyoustartedout.Thefactis,mostentrepreneursarefarmorelikelytoraisetoolittleratherthan
too much money. Nobel Prize–winning economist Daniel Kahneman and hislongtime collaborator, the late Amos Tversky, described this generalphenomenonwhentheywroteabout the“planningfallacy” in their1979paper“IntuitivePrediction:BiasesandCorrectiveProcedures.”
The planning fallacy is that youmake a plan, which is usually abest-case scenario.Thenyouassume that theoutcomewill followyourplan,evenwhenyoushouldknowbetter.
Almost every entrepreneur I’ve ever worked with falls prey to the planningfallacy,especiallyfirst-timeentrepreneurs!Having“extra”capitalgivesyouacushionforwhenoutcomesdonotinfact
followyourplan.Moreover,itincreasesyouroptionality—ifyouneedtoinvestin growth, you can do much more without having to go through the time-consumingprocessofraisinganotherround.AsMariamNaficy,CEOofMinted,toldme, “Act like you’ve got half the amount you have in the bank becauseyou’ve got to factor in all the failures and all the optimizations that kill greatentrepreneursandbusinessesallthetime.Bothofusknowsomanypeoplewhohadgoodideasandwereontherighttrack,butjustranoutofmoney.”AtbothPayPalandLinkedIn,weraisedlargefinancingroundsrightbeforea
marketmeltdown (2000, 2008), andwe surewere gladwedid. In the case ofPayPal,thatmoneyallowedustokeepgrowingduringthedot-combust;withoutit,wewouldn’thavemadeit toourIPO.In thecaseofLinkedIn, thesituationwasn’t as dire, but I realized that the value of the optionality from additionalfundingfaroutweighedthepotentialnegativesofequitydilution.Evenifthemoneydoesn’tprovetobenecessary,amajorfinancingroundcan
alsohavepositivesignalingeffects—ithelpsconvincetherestoftheworldthatyour company is likely to emerge as the market leader, and can discourageinvestorsfrombackingadditionalcompetitors.Mostblitzscalingstart-upshaveahighburnrate.Thisisbecausethedriversof
growth, suchas sales andmarketing,often require significant investments thatexceed the cashcoming fromproduct sales. It usually takes a lotofmoney tomakeakillercompany,whichiswhywehaveventurecapitalists!Butwhile itmaymakesense toburncash inorder togrow(andfinance the
difference with capital you raise from investors), you should make thisinvestment with long-term profitability in mind. If the unit economics arepositiveinthelongrun,andcapitalisavailableatlowcost,thenitmakessenseto take in investment capital to fuel rapid growth. The company won’t beprofitable in the short term, but it is building a customer base that will drivelong-termvalueintheformofgreaterrevenuesandprofitsinthefar-offfuture.Fortechnologystart-ups,theamountofmoneyyouneedtoraisewilltendto
be a function of two primary factors: people costs and the cost of outboundcustomeracquisition.Thegoodnewsisthatthesecostsarelargelypredictable,whichgivesyouachancetoact thoughtfullyrather thanjustreact.Theclassicruleof thumbinSiliconValley is toraiseenoughcashforeighteen to twenty-fourmonthsofoperations.This isbecauseitusuallytakesaboutsixmonthstoraiseyournext roundofventurecapital,whichmeansunlessyouhaveat leasteighteen months of “runway,” you’ll have less than a year to make enoughprogress to convince venture capitalists that you’ve justified another round ofinvestment.Thisisimportantbecauseallfinancingeventsarebetterplayedasalonggame
thanashortgame.We’renotthefirsttoobservethatinvestorsalwaysprefertogive theirmoney to someonewho doesn’t need it. Few things trigger amorepositivePavlovianresponsefromaventurecapitalistthanthewords“Wedon’tneedtoraiseanymoney.”Unfortunately,thisiseasiersaidthandone.When you are blitzscaling your start-up, growth is so rapid and your
organization is pushing its limits in so many ways that multiple things arealwaysbreaking. It’s tempting tofix these thingswithmoney,butyouhave toresistthattemptation.Onlyspendmoneytofixthingsthatareonthecriticalpathtoreachthenextphaseofscale;everythingelsecanwait.AsIdescribedearlier,atPayPalwedeliberatelyavoidedspendingmoneyoncustomerservicebecausewe knew itwasn’t a critical path. Themore you can keep juggling and defer
spending,themorelikelyyou’llbeabletoraisemoneywithoutthepressureofashortrunway.Remember,startingacompanyis likejumpingoffacliffandassemblingan
airplaneon thewaydown. Ifyou runoutofmoney for the fuelandpartsyouneedtogetairborne,noonewillevergettofindouthowefficientlyyouspentitalongtheway!
RULE#9:EVOLVEYOURCULTURE
Nearly all founders, business gurus, and academics agree that organizationalcultureisimportant.Whiletherearealotofinefficienciesyoucantolerateandfiresyoucanletburnduringyourblitzscalingjourney,ignoringyourcultureisnotanoption.BrianCheskyofAirbnbdefinesculture inasimpleandconciseway:“asharedwayofdoingthings.”Clearlydefiningthewayanorganizationdoesthingsmatters,becauseblitzscalingrequiresaggressive,focusedaction,andunclear,hazyculturesgetinthewayofactuallyimplementingstrategy.NetflixcofounderandCEOReedHastingstoldme,“Weakculturesarediffuse;peopleactdifferently,anddon’tunderstandeachother,anditbecomespolitical.”MarkZuckerbergandSherylSandberghavedonemanywonderful thingsat
Facebook, and one of them is building a unified culture that is devoted toaggressiveexperimentationanddata-drivendecisionmaking,assummarizedbyMark’soriginalmotto“Move fast andbreak things.”Facebook’s culturehelpsemployeesunderstandthattheyshouldn’tbeafraidtotrythingsthatmightfail.ThisallowsFacebook tomovefaster,and tomoveonfromfailedexperimentsquickly.Imagine if someone asked a random employee from your start-up the
followingquestions:
Whatisyourorganizationtryingtodo?Howareyoutryingtoachievethosegoals?Whatacceptablerisksareyouincurringtoachievethosegoalsmore
quickly?Whenyouhavetotradeoffcertainvalues,whichonestakepriority?Whatkindofbehaviordoyouhire,promote,orfirefor?
Wouldshebeabletoanswerthosequestions?Ifyouaskedanotheremployee,wouldhegivethesameanswers?Whenorganizationshavestrongcultures,theiremployeesgiveconsistentanswersandactaccordingly.Astrongcommitmenttoaculturewillsometimesmeanpassingonhiring“A
players” who don’t fit that culture. At PayPal, for example, Max Levchininstituted a problem-solving test as part of the hiring process for joining ourengineeringgroup.Hewantedaculturethatwasfocusedonsolvingbig-pictureproblems,notsimplywritinggoodcode.Ifapersonwasagreatprogrammerbutdidn’t have a problem-solving orientation, we didn’t hire him or her. AtLinkedIn, we tried to recruit people who were hardworking but also family-oriented.Ourfoundingteammembershadfamilies,andwewantedtoestablishthenormthatemployeescouldgohometohavedinnerwiththeirfamilies(andthenworkremotelylaterintheevening).Candidateswhobelievedthatastart-upneeded everyone at the office until ten o’clock every night would inevitablyfrustrate colleagues and themselves, so they were screened out. Conversely,candidateswhowantedtoworkanine-to-fivejobwouldalsobescreenedout,nomatterhowtalented.Culture is critical because it influences how people act in the absence of
specificdirectivesandrules,orwhenthoserulesreachtheirbreakingpoint.Inanotoriousexamplefrom2017,actingattherequestofUnitedAirlines,ChicagoDepartmentofAviationemployeesforciblydraggedpassengerDavidDaooffanoverbookedflight,breakinghisnose,knockingouttwoofhisteeth,andgivinghim a significant concussion in the process. The next morning, United CEOOscarMunozsentaratherperplexinge-mailtoUnitedAirlinesemployees.
Our employees followed established procedures for dealing withsituationslikethis.WhileIdeeplyregretthissituationarose,Ialsoemphaticallystandbehindallofyou,andIwant tocommendyouforcontinuingtogoaboveandbeyondtoensureweflyright.I do, however, believe there are lessonswe can learn from this
experience, and we are taking a close look at the circumstancessurrounding this incident. Treating our customers and each otherwithrespectanddignityisatthecoreofwhoweare,andwemustalwaysrememberthisnomatterhowchallengingthesituation.
TheDavidDao incident is a classic example of how a poor articulation of
companyvaluescanweakentheculture.TheemployeesonthegroundbelievedtheyneededtobumppassengersfromtheflightsothatUnitedcouldgetanotherflightcrewto theirplane(i.e.,“flyingright”)and thatmeetingmetricssuchason-time departures and flight cancellations was more important than treatingcustomerswith “respect anddignity” (whichmost of uswould agreedoesnotincludebreakingtheirnosesandknockingouttheirteeth).Incontrast,SouthwestAirlinesisnotonlyclearaboutitscompanyvaluesbut
makesthemtheemphasisofhiringandmanagement.Thementalityisn’t:“We’llknowitwhenweseeit.”Instead,itis:“Doesthispersonalreadylivethewaywedo?” The company uses behavioral interview questions to determine whethercandidatesareaculturalfit.Forexample,todeterminesomeone’sabilitytobeaselflessteamplayer,theymightaskhertodescribeatimewhenshewentaboveandbeyondtohelpacoworkersucceed.Theairlineacknowledgesthatcertainpositionscallforspecificskillsets.As
Southwestputsit,“We’renotgoingtohireapilotwhohasagreatattitudebutcan’tflyaplane!”But,whenitcomesdowntotwoequallyqualifiedcandidates,the one who lives Southwest’s values receives the offer. And, even whenSouthwestfindsaqualifiedcandidatewhodoesn’thavetherightvalues,itwillkeeplookinguntilitfindssomeonewhodoes—nomatterhowlongthejobhasgoneunfilled.Southwest’s development andpromotion practices are also explicitly tied to
company values. In performance reviews, employees are assessed not just onresultsbutalsoonhowtheygotthoseresults; infact,peopleareactuallyratedon things like their “warrior spirit,” “servant’s heart,” and “fun-LUVingattitude.”In other words, culture isn’t just manifested as a mission statement on
Southwest’swebsite;it’swovenintotheairline’sprocessesandpractices.In both cases, culture has real business implications. In the case of United
Airlines, a culture without strong values led to a disastrous public relationsfiasco. For Southwest Airlines, the value of employees’ “servant’s heart” and“fun-LUVing attitude” shows up in the concrete metrics of customersatisfaction.Even thoughSouthwest’sperformanceonmetrics suchason-timearrival is decent to above average, the airline consistently has the fewestcustomercomplaintsperpassengercarried.OrganizationalculturehasplayedakeyroleintheriseofSiliconValley.Most
oftheiconiccompaniesthathaveshapedanddefinedthetechnologyindustry—
Hewlett-Packard, Intel, Apple, Google, Facebook—are known for theirdistinctivecultures,regardlessoftheirera.Thesamecanbesaidformorerecentstart-upmarketleaderslikeAirbnbandSalesforce.com.Typically,thecreditfortheseculturesgoestothefounders.BillHewlettand
DavidPackardaresynonymouswiththeHPWay.BobNoyce,GordonMoore,andAndyGrovearereferredtoastheIntelTrinity.SteveJobs,LarryPageandSergeyBrin,andMarkZuckerbergareseenasthesourcesofApple’s,Google’s,andFacebook’scultures.Yetwhilethepersonalitiesofthefoundingteamplayacriticalroleindefininganorganization’sculture,itismoreaccuratetosaythatan organization’s culture emerges over time based on the actions of manypeople,notjustthefounders.Theprimary culture of anorganization typicallyoriginates in the functional
area that ismost critical to the successof the company.Earlyon, engineeringpredominated, and engineering culture formed the basis of things like theHPWay.Asthe technologyindustrymatured,salesassumedagreater importance,andsales-orientedculturesaroseatcompanieslikeOracleandCisco.Companiestoday might also have a product culture, design culture, marketing culture,financeculture,orevenanoperationsculture.Anyandalloftheseculturescanbe successful, but you should focus on whichever function is key for yourorganization to succeed. In addition to the role this choice plays in thecompany’svalues,theexecutiveinchargeofthefunctionalareathatdrivestheculturealsotendstobethemostlikelysuccessortotheCEO.The development of organizational culture is intimately intertwined with
branding.Culture iscentral to thestory thatwetellourselvesandothersaboutwhoweareandourplaceintheworld,aswellasthestoriesotherstellaboutus!Youcanlistallthevaluesyouwantonyourwebsite,buttheonlywaythattheybecome an integral part of your culture is if you make them a part of yourstrategy, andmake sure that you and others can tell stories (backedwith hardevidenceandconcretedetails)aboutlivinguptothem.Sohowdoyoudevelopastrongcultureatyourorganization? Ibelieve that
the best approach represents amiddle ground between hoping that the cultureevolvesorganicallythroughbenignneglectandtryingtodefineacomprehensiveculture up front. The former approach risks developing aweak culture or onethatdoesn’tfitthecompany’sneeds;thelattermaybetoorigidandinflexible.Mostculturesbegin to formorganically.Aswe’vediscussedpreviously, the
founders of the organization have a major influence on the culture, simply
becauseofwhotheyare.Ifafounderbelievesthatcertainbeliefsandpracticesare fundamental keys to winning, those beliefs and practices tend to betransmitted to thepeoplewhoworkcloselywithhimorher.Thismightoccurviafilteringduringthehiringprocess,asaresultofworkingcloselytogether,orboth. For example, Larry Page of Google is a technologist with a strongacademic background. As a result, Google developed a technology-oriented,academicculturethatstronglyresembledStanford’sgraduateComputerScienceDepartment.For example,Googleengineers sat in four-personofficesbecausethat is how Stanford organized graduate students’ offices. Google hired EricSchmidt to bring in stronger business experience, knowing that Eric’s ownacademicbackground(heearnedaPhDinElectricalEngineeringandComputerScience [EECS] at UCBerkeley) would allow him to effectively synchronizeculturallywithLarryandcofounderSergeyBrin.In blitzscaling companies, culture becomes increasingly important—and
increasinglydifficult tomaintain—astheorganizationgrows.In thebeginning,the bond that employees form in the early days of the company can be apowerful force in shaping the culture, but this getsmore challenging asmorepeoplecomeonboardandspontaneousinteractionsgivewaytomoreformalizedstructures.Transmittingcultureorganically requiresbothpersonal interactionand time.
ThisosmosisworksduringtheFamilyandevenTribestagesofblitzscaling,butbreaksdownatlaterstages.Ifthefoundersdon’thavepersonalinteractionswithall employees, or if those interactions are brief and sporadic, osmosis doesn’twork.Andwhenacompanyisdoublingortriplinginsizeeveryyear,evenbriefandsporadicpersonalinteractionswiththefoundersareabest-casescenario!BythetimeanorganizationreachestheVillagestageofblitzscaling(atleast
one hundred employees), the mesh of person-to-person interaction isinsufficient, especiallywhen culture needs to be synchronized acrossmultipleoffices.DrewHoustonmakes sure that all Dropbox employees are aware that they
needtohelpre-createtheculture.“Wetellpeople,‘Youmighthavejustjoinedlast week, but sooner or later, you’ll be an old-school Dropboxer too. Soremember the things you like about this place now, because it’ll be yourresponsibilitytomakesurethosethingsstickaround.’ ”It’snotalwayseasytomovefromorganictodeliberateculturaltransmission.
ReedHastings’sexperienceistypical.“Whenwe[Netflix]wentpublic,wehad
150people,”Reedtoldme.“Peoplewereworriedthatnowthatwewerepublic,everythingwouldgotoshit—we’dputinalotofprocess,andstoptakingrisks.Whatwe’vedoneistopromoteemployeefreedom.Ifyouwanttooperatewithveryfewrules,youneed tosetcontext.”Yetwhileearlyemployeesoften fearthat deliberate cultural development will bring bureaucracy, as Reed argues,cultureisactuallyasubstituteforbureaucracyandrules.Thestrongeryoumakeyourculture,thelessyou’llhavetobindpeople’sbehaviorwithrigiddirectives.The two key levers of deliberate cultural transmission are communications
and peoplemanagement. Communications are important because they providefounders with a direct channel to all employees. This can take many forms,rangingfromformalin-personmeetingstoelectroniccommunicationstothingsasseeminglyneutralasofficelayoutanddesign.Airbnb,forexample,employsawiderangeofchannelstomaximizecultural
transmission. Theweekly e-mail cofounder Brian Chesky sends to all Airbnbemployeesisapowerfulone.“Youhavetocontinuetorepeatthings”BriantoldourclassatStanford.“Cultureisaboutrepeating,overandoveragain,thethingsthat reallymatter foryour company.”Airbnb reinforces theseverbalmessageswithvisualimpactaswell.BrianhiredanartistfromPixartocreateastoryboardoftheentireexperienceofanAirbnbguest,fromstarttofinish,emphasizingthecustomer-centereddesignthinkingthatisahallmarkofitsculture.EvenAirbnbconferenceroomstellastory;eachoneisareplicaofaroomthat’savailableforrentontheservice.EverytimeAirbnbteammembersholdameetinginoneofthoserooms,theyareremindedofhowguestsfeelwhentheystaythere.At Amazon, Jeff Bezos famously bans PowerPoint decks and insists on
writtenmemos,whicharereadinsilenceatthebeginningofeachmeeting.Thismemo policy is one of the ways that Amazon encourages a culture of truthtelling.Memoshavetobespecificandcomprehensive,andthosewhoreadthememoshavetorespondinkindratherthansimplysitthroughsomebroadbulletpoints on a PowerPoint deck and nod vague agreement. Bezos believes thatmemosencouragesmarterquestionsanddeeper thinking.Plus,becausethey’reself-contained(rather thanrequiringapersontopresentadeck), theyaremoreeasilydistributedandconsumedbyawiderpopulationwithinAmazon.The late Steve Jobs used architecture as a core part of his deliberate
communications strategy at Pixar. He designed Pixar headquarters so that thefrontdoors,mainstairs,maintheater,andscreeningroomsallledtotheatrium,which contained the café and mailboxes, ensuring that employees from all
departments and specialties would see people from other groups on a regularbasis, thus reinforcing Pixar’s collaborative, inclusive culture. In WalterIsaacson’sbiographyofSteveJobs,JohnLasseter,Pixar’schiefcreativeofficersays,“Steve’stheoryworkedfromdayone.IkeptrunningintopeopleIhadn’tseen for months. I’ve never seen a building that promoted collaboration andcreativityaswellasthisone.”
Builda“ShipofTheseus”
The other main lever for cultural development is the organization’s peoplemanagement practices. After all, the strongest influences on organizationalcultureareoftenwhoyouhire,promote,andfire.WhenhevisitedourBlitzscalingclassatStanford,EricSchmidtsharedhow
Google’shiringstrategyshapeditsculture.“Thepeoplethatyouhiremakeyourculture,”Ericsaid.“We’dhirepeoplewhowerespecialinsomeway.Youdon’thire generic people—you hire people who have had stress and achievement.”Culture is a key part of the hiring process atAirbnb too. Each candidate alsogoes through a values interview, conducted by anAirbnb employeewho isn’tthat candidate’s hiring manager. This ensures that values are consideredindependently of howmuch the organization needs that candidate’s particularjobskills.Whenthebusiness isgrowingthatquickly, it’susuallydesperateforbodies,
and it can be tempting to simply paywhatever is necessary to get employeesthroughthedoor.Theproblemisthatyouenduphiringmercenariesratherthanmissionaries.Andifyou’retriplingthecompany’ssizeeachyear,youcanshiftyour company from a majority-missionary culture to a majority-mercenarycultureinasingleyear.Another side effect of rapid growth is that many, if not a majority of,
employeeswillbereportingtoinexperiencedmanagers.Asystematicapproachlike that laid out in The Alliance can help those managers better align thepersonalvaluesandmissionsofemployeeswiththecompanyculture.Itcanbetoughtoprioritizeculturewhenotherfiresareburning.Duringthe
Family and even Tribe stages of blitzscaling, HR often doesn’t exist as aseparate department and might be outsourced to a professional employerorganization like TriNet, or left as a part-time job that an office manager oradministrative assistant handles as a secondary priority. As a result, the
unconscious habits and patterns of early employees often form and crystallizeintoanorganizationalculturewithoutanyexperiencedoversight.Andevenafterthe company adds an HR function, its first priority is usually to hire moreemployeesasquicklyaspossiblerather thanfocusoncultureandvalues.IfanorganizationwantstomakecultureanHRpriority,thefoundersandleadershipneedtomakesurethattheHRteamisgiventhetimeandresourcestodoso,andthatnewhiresaremanaged,evaluated,andrewardedaccordingly.Thesemechanismsneedtocontinuetoevolveas thecompanygrowsandits
needsevolve.ReedHastingsandNetflixarewell-knownfortheNetflixCultureDeck, a one-hundred-plus slide presentation that explains Netflix’s high-performanceculture.ReedandPattyMcCordcreated theNetflixCultureDecktohelpfilteroutjobcandidateswhowouldn’twanttoparticipateintheNetflixculture.But the deck isn’t carved in stone;Netflix continues to revise it on aregularbasis.Oneofthereasonsforevolvingyourcultureisthe“ShipofTheseus”paradox.
TheancienthistorianPlutarchcoinedtheterminreferencetotheshiponwhichthemythicalheroTheseusreturnedtoAthensafterslayingtheMinotaur.Asthelegendgoes,theAthenianshadpreservedthefamousvesselbyreplacingbrokenpartswithnewwood,untilatlastnoneoftheoriginalwoodremained.Plutarchreported that philosophers argued strenuously, and without resolution, overwhethertheshipofreplacementpartswasstilltheShipofTheseus.(Amusingly,the philosopher Thomas Hobbes complicated matters by asking what wouldhappeniftheoriginalwoodpartswerepreservedafterbeingreplaced,andwerethenusedtobuildasecondship!)Allcompaniesare like theShipofTheseus.Employees join,stayforoneor
moretoursofduty,andleave,onlytobereplacedbynewemployees.Astable,low-growth company might persist for decades or longer, slowly replacingemployees but remaining the same size and retaining a strong sense ofcontinuity.Inotherwords,the“planks”oftheshipremainessentiallyunchangedfromdecadetodecade.Incontrast,ablitzscalingcompanylikeFacebookmightgrowfromtheFamilytoNationstageinasingledecade,doublingortriplinginsize each year, so that the employees who made up the entire ship on NewYear’sDayendupasasmallminoritybythefollowingNewYear’sEve.Atthesame time, as we discussed in “HireMs. Right Now, NotMs. Right,” manyearly employees are likely to leave at some point along the journey, whichmeans even fewer of the original “planks” are still part of the ship.Yet thesechanges are a necessary part of blitzscaling: you need new people with new
skillsasyougrow.The people, the product, and the offices of a company can, will, andmust
changeasitblitzscales.Cultureisoneofthefewmechanismsthatallowtheshipto retain its essential identity. Culture is what helps Apple retain its “Apple-ness”withSteveJobsgone,andIntuitretainits,well,“Intuit(ive)-ness”evenasitshiftedfromsellingpackagedpersonalfinancesoftwaretoprovidingacloudaccountingsuite.Organizationalculturehasbecomesuchahottopicinthisageofblitzscalingbecausecultureismoreimportantwhenthereisrapidgrowthandchange,ratherthanstabilityandstasis.Youhavetowalkafinelineasyouevolveyourculture—evolveittooslowly,
and it will hold you back from adapting to new businesses and the changingworldaroundyou.Evolveittooquickly,andtheShipofTheseusillusionbreaksdown,andpeoplenolongerfeelliketheybelong.In thewords of theDutch historian JohanHuizinga, “Ifwe are to preserve
culture,wemustcontinuetocreateit.”
ALackofDiversityandOtherCulturalPitfalls
Given the popular emphasis on company culture, it’s also important to offersome thoughts on the potential pitfalls that can come with trying to build astrongculture.First, there can be a fine line between a strong culture and a cult. By
definition,cultureisnarrowingtosomedegree.Buildingcultureintoyourhiringprocessesmeans that you’re excluding people by design, and you have to becarefulnot to restrictyourhiring to thepointof totalhomogeneity.Successfulorganizationsneedacombinationofconformityanddiversity.Therightkindofsameness(e.g.,smart,driven,intelligent,hardworking,mission-driven)cangiveacompanyanedge,aswascertainlythecaseatPayPal.Buttoomuchsamenesscanresultingroupthink,bias,andstagnation.Fartoomanycompaniesmisinterpretwhatitmeanstohirefor“culturalfit.”
Inmanycases,thisleadstoteamsthatareheavyonyoung,Caucasianmenwhowenttoashort listofelitecolleges,whichhinderstheorganization’sabilitytoinnovateorserveabroadermarket.Butevenwithoutsuchproblematicpractices,hiringfor“fit”shouldnotmeanasking,“Doyoufitintothisbox?”Forexample,manystart-upshaveaworkculturewhereemployeescome to
the office after 10 a.m.,work late, and spend evenings socializing together atbars.Inotherwords,anextendedversionofcollege!Ifyourstart-uphassuchaculture,youmightavoidhiringemployeeswhowanttocomeinearlyandleavebefore6p.m.,orwhorarelysocializeorgooutatnight.Thatmightbegoodfor“cultural fit,” but it also means that you won’t hire people who don’t drinkalcoholforreligiousorotherreasons,youwon’thireemployeeswhoareparents,andyouprobablywon’tevenhireemployeeswhoaremarried(oratleastthosewhowishtoremainmarried).Ratherthanhirepeoplewho“fit”yourcultureinsuperficialways—whetherbasedongenderorraceoralmamater—hirepeoplewhoareadditivetoyourculture.WhenBelindaJohnsonjoinedAirbnbin2011,shebroughtaverydifferentbackgroundandexperiencetotheyoungcompany.Thefounderswereintheirtwenties;Belindahadbeenalawyerwhentheywerestill in diapers, andhad spent a dozenyears as an executive atYahoo! Itwasprecisely these differences that helped both Belinda and Airbnb succeed as ateam.BrianCheskycallsherAirbnb’s“SecretaryofState,”andherdiplomacyandknowledgehelpedthefirmdevelopproductiverelationshipswithregulatorsandmunicipalities.Newhiresareanopportunitytorefineyourcultureandaddtoitscapabilities.Theyshouldbecompatiblewithyourcurrentculturebutalsobringelementsthathelpchangeitforthebetter.Theartisinfindingtransplantsthattheorganization’sexisting“immunesystem”doesn’treject.Blitzscaling companies areparticularly susceptible tobuilding a culture that
lacks diversity because of their relentless emphasis on speed. The fastest andeasiestwaytohireisusuallytoaskemployeestorefertheirfriends.Buthiringbasedonhomophily(“birdsofafeatherflocktogether”)almostinevitablyleadstohomogeneity.Justasstart-upsincur“technicaldebt”bytakingshortcutswiththeircode, theycan incur“diversitydebt”by takingshortcutswith theirhiringpractices.This diversity debt is a serious problem for individual companies, and for
societyasawhole.Homogeneityharmscompaniesbecausegroupthinkreducesthe resilience and adaptability of the companies, and it harms society if thosemanyopportunitiesblitzscalingprovidesaren’tfullyopentoallqualifiedpeople,whatevertheirgender,sexuality,religion,orancestry.Oneof theugliestmanifestationsof theseproblems is the cultureof sexism
andsexualharassmentthathasbeenuncoveredatvariouscompanies.Inalmosteverycase,theseproblemsarosebecauseoneormoreemployeeswhobelongedtoademographicgroupthatrepresentedtheoverwhelmingmajorityheldpoweroveremployeeswhowereintheminority.Manytimes,executivesabusedtheir
power, setting a disgraceful example for other employees. This is absolutelyunacceptable,andbegsforaction.In2017,forexample,IcalledfortheDecencyPledge to try to address the serious problems in the venture capital industryaroundmenabusingtheirpowerandpositiontoharmwomen(andsomemen).Hopefully,mostblitzscalingcompanieswillneverreachthepointwheretheir
cultures tolerate suchbadbehavior, but thebestway to ensure this is tobuildinclusive cultures from the very beginning. This is one area in which it’sinsufficienttoallowculturestoevolveorganically.EvenattheFamilystage,acompanyshouldbeexplicitaboutdiversity,andstateinwritingthatitstrivestobe inclusive in terms of gender, sexuality, religion, ancestry, and age.And, itshouldmakediversity a priority in hiring startingwith its first ten employees,andespeciallyincorefunctionslikeproduct,engineering,andmarketing.In theTribeandVillagestages, the rateatwhichyouareaddingemployees
will require a more systematic approach to diversity. We recommendimplementingatleastthreekeypolicies:First,measureyourdemographicsandmakethat informationtransparentandavailable,bothinternallyandexternally.Aswithanymetric,youcan’tmanagewhatyoudon’tmeasure.Second,institutethe equivalent of theNationalFootballLeague’sRooneyRule,which requiresNFLteamstointerview(butnotnecessarilyhire)atleastoneminoritycandidatefor any senior football operations position. And third, tie at least part ofexecutivecompensationtothecompany’sprogresstowarditsdiversitygoals.Inanidealuniverse,allcompanieswouldbuilddiverseworkforcesfromthe
start. But the more people a company employs, the more important diversitybecomes.Don’twaittomakediversitygoalsapriority.It’smuchhardertoshiftfroma“brogrammer”haventoa truly inclusiveculturewhenyou’rea10,000-personcompany.Anotherpitfall isculturalhypocrisy.Ifyoupreachthegospelofyourstrong
culture,youhavetoliveuptoit,oryou’llbedoingmoreharmthangood.Whenyou talk the talk but don’t walk the walk, employees will recognize thehypocrisy.Credibilityhastobeearned,notsimplyasserted.Thisisparticularlytrueoffounders,whotypicallyhavethemoralauthoritywithinastart-up,andofnonfounder CEOs, whose position magnifies the impact of their words andactions.FoundersandCEOsareculturalrolemodels;iftheydon’texemplifytheculture,itwillinevitablyweaken.
THENEVER-ENDINGNEEDFORCHANGE
What do the eight key transitions and nine counterintuitive rules have incommon? They reflect the fact that when you’re blitzscaling, the need forchangeneverstops.Justwhenyou’vemanagedakeytransitionorsuccessfullyappliedacounterintuitiverule,thegameboardchanges,andyouhavetodoitallagain.Nomarket remainsvaluable forever,whichmeans that evencompanies that
successfullyreachtheNationstagebydominatinganimportantmarkethavetokeep searching for the nextmarket inwhich to blitzscale.Every exciting newtechnology or market that once supported massive wealth creation eventuallybecomesa stable,boring industry.Atvariouspoints inhistory, thecargoship,railroad, and automobile industries spawned companies and innovations thatchangedtheworldandmadegenerationalfortunes.Today,they’relargelysleepybackwaters (though occasionally companies like Tesla manage to revitalizethem), a fate that is still better than irrelevance but unlikely to hold manyexcitingopportunitiesforlarge-scalegrowth.Thesamepatternhasplayedoutonasmallerscale inSiliconValley,where
markets for Dynamic Random-Access Memory (DRAM), hard drives, andpersonalcomputersallowedcompanieslikeIntel,Seagate,andCompaqtogrowtomassivevaluebeforebecominglow-margincommodities.(Intelcontinuedtogrow, thanks to its shift to high-margin CPUs, while Seagate and Compaqlanguished,andinthecaseofCompaq,endedupbeingacquiredandvanishing.)The best entrepreneurs and companies use successful blitzscaling in one
markettojumpintoanother.InteljumpedfromDRAMstomicroprocessors,androde a secondwave to even greater heights.Microsoft used its dominance inoperating systems to develop the even more dominant Microsoft Officeplatform.Amazon’se-tailblitzscalingallowedit tobecomethe leader incloudcomputing with AmazonWeb Services. Perhaps Facebook will do somethingsimilarwithVR.The never-ending need for change should fill youwith both fear and hope.
Fear,becauseyoucanneverrestorstandstill.Hope,becausenewmarketsarealways emerging, giving everyone, from Silicon Valley to Shanghai (andeverywhereinbetween),theopportunitytobuildnothingshortofanewrocketship.InLewisCarroll’s classic bookThrough theLooking-Glass, theRedQueen
tellsAlice,“Now,here,yousee,ittakesalltherunningyoucando,tokeepinthesameplace.Ifyouwanttogetsomewhereelse,youmustrunatleasttwiceas
fastasthat!”Sometimesblitzscalingacompanymightfeelabitlikerunningashardasyoucansimplytoendupinthesameplace.ButthedifferencebetweenourworldandtheRedQueen’sisthatblitzscalingisaracetobuildthingsthatmake the world a better place. Whether your new market will be machinelearning, or a new kind ofwireless computing, or something that hasn’t beeninventedyet,there’sawordfortheby-productofblitzscaling:“progress.”
PARTV
TheBroaderLandscapeofBlitzscaling
While many of the examples you’ve encountered throughout this book havebeenSiliconValley technologycompanies, theprinciplesofblitzscaling applyfar beyond this specific domain. In this section, we’ll focus on the broaderlandscape of blitzscaling, including how it works in other geographies andindustries,andwhatitmeansforthefutureofourglobaleconomy.Becauseitissuchanimportant topic,we’lldevoteanentiresectionof thischaptersolelytoexaminingblitzscalinginChina.
BLITZSCALINGBEYONDHIGHTECH
Whileblitzscaling is probablymost applicable tohigh tech, its techniques canbenefit any industry in which opportunities can demonstrate strong growthfactors (market size, distribution, gross margins, and network effects) andovercome the growth limiters (lack of product/market fit and operationalscalability).Forexample,considertheSpanishclothingretailerZara.Fewindustriescould
seem farther away from the world of Internet companies like Google andFacebook. Yet while Zara’s expansion took longer to get there (the companywasfoundedin1975,which,coincidentally,isthesameyearthatMicrosoftwasfounded), its scale and dominance in its industry rival those of its high-techequivalents, and have made its founder, Amancio Ortega, the world’s third-richestman(behindJeffBezosandBillGates,butaheadofWarrenBuffett).Zara plays in an enormousmarket; global apparel sales in 2016were over
$1.4 trillion, and even though Zara’s gross margin reached a ten-year low in
2017, it still stood at a robust 57 percent (versus 61 percent for Google, and35percentforAmazon).Itsglobalnetworkofstoresgivesitbroaddistribution,and while apparel doesn’t offer any strong network effects, clothing doescommandafairamountofconsumerloyalty,enablingZaratolockinadegreeoflong-termadvantage.What’s even more important, however, is that Zara actually uses the
techniquesofblitzscalingtorunitsbusiness.Speedis thefoundationofZara’s“fast fashion” business strategy, which, for decades, can be summarized in asinglesentence:“Givecustomerswhat theywantandget it tothemfaster thananyoneelse.”EveryaspectofZara’sbusinessisorganizedaroundachievingthatspeed.The
resultsareimpressive:Zaratakesonlytwoweekstodevelopanewproductandget it into stores—the industry average is sixmonths—and launches over tenthousand new designs per year, a rate several times that of competitors likeH&MandGap.Zaraholds just sixdaysof inventory,while rivalH&Mholdsnearlytentimesasmuch.Inthe1970s,OrtegaestablishedarulethatZarahadtofulfillapparelorders fromitsstores in less thanforty-eighthours.Today,Zarastillfollowsthatrule,eventhoughithasexpandedfromalocalSpanishretailerintoaglobalempirewithstoresinAfricaandAsia.To achieve these results, Zara has found a way to balance addressing the
growth limiter of operational scalability while following the counterintuitiverules toembracechaosanddo things thatdon’tscale.Given itsmassivescale,youwouldexpectZaratoturntoChinaforhelpwithimprovingitsmargins,intheway thatApple didwith the iPhone. But unlike its competitors, Zara stillmanufacturesmostof itsclothing inSpain.Thanks to its financialmight,Zarawas able to build fourteen highly automated factories in Spain, where robotscreate“greigegoods”—newlymanufacturedclothesthathaven’tbeenwhitenedand dyed.Zara then uses a partner network ofmore than three hundred smallshopsinSpainandPortugaltoprocessthosegreigegoodsintofinishedapparel.WhilethelaborcostsmightbehigherthaninChina,andthusless“efficient,”thepayoffisincredibleresponsivenessandspeed.Responsive manufacturing is critical to Zara’s business model. Apparel is
designed by small teams in Zara’s design center, where designers work withpatternmakersandcommercialsalesspecialists.FeedbackcomesindailyfromZara’s storemanagers. That feedback is analyzed by the sales specialists andthenpresentedtothedesignersandpatternmakers,whostartsketchingdesigns
onthespotandcreate,onaverage,anastonishingthreenewitemsperday.Thedesigns are then dispatched to those factories for manufacturing and on topartnershopsforfinishing.Zara’s logistics model continues this preference for responsiveness over
efficiency. Zara products are distributed in small batches,which requiremorefrequent shipments. The logistics costs are higher, but this allows Zara to getclothestoitsstoresinlessthantwenty-fourhoursforEurope,theMiddleEast,andAmerica,andinlessthanforty-eighthoursforAsiaandLatinAmerica.Thisfocusonspeedcomesdirectlyfromthefounderandanimatestheentire
organization.Ina2013profile,Fortunetoldthestoryofhow,stoppedatatrafficlight, Ortega spotted a young motorcyclist wearing a jean jacket covered in1970s-stylepatches.Ortegagrabbedhiscellphone,calledanaide,describedthejacket over the phone, and ordered the aide to get the design into production.Loreto García, the head of Zara’s woman’s trends department, explained theneedtobelightningfastinrespondingtotrendswhenshetoldFortune,“Whatseemsgreattoday,intwoweeksistheworstideaever.”Despiteall thechaos,and the inefficiencyofmanufacturingandshipping in
smallbatches,Zara’sgrossmarginscontinuetoexceedthoseofitscompetitorsH&M (55 percent) and Gap (29 percent). That’s because all that inefficiencyincurredinthepursuitofspeedallowsZaratoavoidoneofthebiggestdragsongrossmarginforalmostanyapparelcompany—overstockofdesignsthatfailedtosell.Ortegadevised thismodelwhenhewassixteenyearsold—don’torderinventoryandhopeitsells;instead,figureoutwhatpeoplewant,andthenmakeit.Anotherexampleofhowblitzscalingappliestoacompletelydifferenttypeof
businessistherapidriseoftheshaleoilandnaturalgasindustryintheUnitedStates during the 2000s. The energy sector scores well on the growth factorswe’vedefined.Oil andgas is an enormous, high-margin industry,with a veryefficientdistributionsystem.Andwhiletheshaleindustrydoesn’tfeaturemanynetwork effects, it has its own source of powerful long-term competitiveadvantage. In the energy industry, rather than buy land outright, the usualpracticeistoleasethedrillingrightsforninety-nineyearsforacombinationofguaranteedleasepaymentsandroyalties.Thismeansthatleasingtherightlandistantamounttoholdinganunbreakablemonopolyontheoilandgasunderneaththatland,atleastforthetermofthelease.Blitzscaling allowed shale companies to grow at an amazing rate. In 2002,
leadingshaleplayerChesapeakeEnergyreportedrevenuesof$738million.Justfouryearslater,Chesapeakereportedrevenuesof$7.3billionandwasaddedtotheS&P500.That’sanorderofmagnitudeofgrowthinthetimeittakestogothroughhighschool.Chesapeake’s cofounders, the lateAubreyMcClendon and his partner Tom
Ward, didn’t have the usual industry background in exploration or refining.Instead ofworking drilling rigs or operating refineries,McClendon andWardwere “landmen,” specialistswhowent out into the field to negotiatemineralrightsleaseswithlandowners.Thisexpertisewouldbekeytotheirblitzscalingeffort.In the late 1990s, the combination of horizontal drilling and improved
hydraulic fracturing techniques (fracking) made extracting hydrocarbons fromshale rock formations economically feasible for the first time. Essentially,energy companies could drill horizontal shafts into rock formations and thenpumphigh-pressureliquidsintothewellstofracturetherockandreleasemoreoil and gas. Because traditional drilling techniques didn’t work on shale rockformations,thelandabovethoseformationshadneverbeenleased,whichmeantthat when frackingmade those hydrocarbons accessible for the first time, themarkettoacquirethosemineralrightswascompletelywideopen.While Chesapeake wasn’t the first company to employ fracking—it was in
factMitchellEnergy,whichbroughtfrackingtotheBarnettShaleformationinTexas in 1997—it combined this technology innovation with McClendon’sbusinessinnovationtobecomethefastest-growingenergycompanyinhistory.Chesapeakemoved faster thananyother company in its industry,deploying
an army of land men to aggressively lease as much land as possible, withinstructions to paywhateverwas necessary,without knowingwhether the gasdeposits would justify the price. Hiring an army of landmen and paying topdollar for leases sight unseen seemed inefficient…until the wells startedproducing.Chesapeake’swillingness to blitzscale paid off as improvements infrackingtechnologymadeitswellsinsanelyprofitable—atfirst.ThecaseofMcClendonandChesapeakealsoillustrates theinherentrisksof
sacrificingefficiency in thenameofhypergrowth.Blitzscalingcan lead tobigwinsandbiglosses,sometimesatthesamecompany.Chesapeakecontinuedtoborrowmoney to lease more land at ever higher prices.McClendon acted asthoughhisblitzscalingstrategywasguaranteedtosucceed,andChesapeakewashithardbytheglobalrecessionof2008.Afterpeakingat$62.40inJune2008,
its stock price has declined sharply, falling as low as $2.61 in early 2016 (in2017,ittradedbetween$4and$8pershare).McClendonhadtakenonalotofriskinhispersonalfinancesaswell,borrowingmoneytobuyChesapeakestock.Amargincallin2008forcedhimtosell94percentofhisChesapeakestockatamassiveloss.McClendonwaseventuallyforcedtostepdownfromhispositionasCEOof
Chesapeakein2013,butremainedaresoluteblitzscaler.Atthetimeofhisdeathin 2016, McClendon was running American Energy Partners, a company hefounded after his departure from Chesapeake, and for which he had raised$15billionfrominvestors.Even if you are not able to fully apply the business model patterns of
software-drivenhightechtoyour industry,acarefulanalysisofgrowthfactorsand growth limiters might just show you where to find an opportunity toblitzscaleandreaptheassociatedrewards.Afterall,ifyoucanblitzscaleT-shirtsandoilwells,youmightverywellbeabletoblitzscaleanybusiness.
BLITZSCALINGWITHINALARGERORGANIZATION
While thehypergrowthofblitzscaling isoftensynonymouswithscrappystart-ups,blitzscalingcantakeplacewithinlarger,establishedorganizationsaswell.None of the growth factors or limiters, and none of the proven patterns orbusiness models, require a business to be an independently owned, privatelyheld, venture-backed corporation. Even if your organization can’t offer stockoptionsthatwillmakeyouremployeesrichifyourblitzscalingissuccessful,youcanandshouldadoptandadapt the lessonsofblitzscaling tohelpyouachieverapidgrowthandfirst-scaleradvantage.Trying to blitzscale within a larger organization has both advantages and
disadvantagesoverdoingsowithinastart-up.It’scriticaltoberealistic—start-upshavesomeinherentadvantageswhenitcomestoblitzscaling.Blitzscalingisallaboutspeedandrisktaking,andwithfarlesstolose,start-upsaremuchmorenimble. Established companies that want to blitzscale need to find majoradvantagestoovercometheirinherentdisadvantagesinspeedandrisktaking.
ADVANTAGE#1:SCALE
Thismight soundobvious, but there are someopportunities that you can onlytackle ifyoualreadyhave the scale that comeswithbeinga large, establishedplayer. For example, Amazon couldn’t have launched AmazonWeb Services(AWS)without achievingmassive scale in its data centers, and becoming theworldleader inmanagingthosedatacenters.Tryingtobuild thatproductfromscratch, without being able to leverage Amazon’s economies of scale andreputationforoperationalexcellence,wouldhavebeennearlyimpossible.Eventoday, AWS’s main competitors come from other scale companies likeMicrosoft,Google,andIBM.Outsideoftech,scalecanbeanevenbiggeradvantage.WhenQuickenLoans
launchedRocketMortgage,whichprovidesmortgagesonlinewithadecisioninless than tenminutes, RocketMortgage was able to tap into Quicken Loans’consumer marketing expertise—including a Super Bowl ad—to acquirecustomers, and the company’s existing financial relationships to fund thosemortgages.Asaresult,initsfirstfullyearofoperation(2016),RocketMortgageoriginated $7 billion in loans, which would rank it in the top thirty of allmortgage lenders in the country if itwas an independent company, andwhichhelped push Quicken Loans’ overall closed loan volume to $96 billion, upsharplyfrom$79billionin2015.Ontheotherhand,ifastart-upcanplaythesamegame,scalemaynotprovide
a significant advantage unless there is a massive difference in scale. Forexample,whenAirbnbwasblitzscaling,itwascompetingwithHomeAway,anestablished player that had much greater scale. However, HomeAway hadachieveditsscaleviaastringoftwenty-oneacquisitions,whichmeantthatallofits acquisitions were running on different technology platforms and servingdifferent clienteles. Indeed, HomeAway’s scale was actually a disadvantage!HomeAway itself was later acquired by Expedia, as part of that company’sresponsetothecompetitivethreatofAirbnb.
ADVANTAGE#2:ITERATION
Another advantage that established companies have is the ability to makemultiple,iterativeblitzscalingattempts.Blitzscalingisariskystrategy,andyoumightnotachievesuccesson the first try.Youneed tohaveenoughcapital tostay in the game.Microsoft was famed for its ability to iterate its way fromknockoff products to market dominance. The first and second versions of
Microsoft Windows were unsuccessful attempts to copy Apple’s Macintoshoperating system; the third version, while still inferior to its inspiration, wasgoodenough,andMicrosoftunleashedamarketingblitzforfollow-upversionssuch as Windows 95 and Windows NT that carried them to dominance.Microsoft later repeated this strategy with its Xbox business, which evolvedfromtheXbox,totheXbox360,totoday’sXboxOne.Toborrowan analogy fromsports, youmayneed to take repeated shotson
goal before scoring. Established players have a much easier time financingmultipleshotsongoal.Noristhisadvantagelimitedtotechnology.Intheshaleoilindustry,financial
wherewithal played a major role in the success of pioneers like ChesapeakeEnergy.ThelateAubreyMcClendontoldRollingStonein2012,“Tobeabletoborrowmoney for ten years and ride out boom-and-bust cycleswas almost asimportantaninsightashorizontaldrilling….Ifsomethingdidn’tworkforalittlebitoftime,wecouldregroupandfindsomethingthatdidwork.”
ADVANTAGE#3:LONGEVITY
Whiletheabilitytoundertakemultipleattemptsatblitzscalingisanadvantage,soistheabilitytobepatientwithasingleattempt.Largecompaniescan(iftheyhavepatientshareholders)havelongertimehorizonsthanstart-ups,whichneedto show immediate results to continue raisingmoney.Google often plays thislonggamewithtechnologiesrangingfromself-drivingcarstoacureforaging.Facebook is alsoplaying the longgamewithOculusRift andVR.Thekey isknowingwhentoscaleup.Microsoft tried toscalesmartphones tooearlywithWindowsCE;asitturnsout,themodernsmartphoneonlybecamepracticalonceMoore’s Law made mobile CPUs powerful enough, and Apple combinedsoftware with capacitive touch screens, Corning’s damage-resistant GorillaGlass,andhigh-volumeChinesemanufacturing.
ADVANTAGE#4:MERGERS&ACQUISITIONS(M&A)
One final advantage that established players have is the ability to useacquisitions to drive blitzscaling. Acquiring a business that is alreadyblitzscalingorhasthepotentialtoblitzscalecantransformanexistingcompany.
Priceline,forexample,bestknownfor“nameyourprice”airfares,executedthisstrategy to perfectionwhen it acquiredBooking.com, allowing it to achieve alastingadvantageinthehotelbookingmarket.ManyAmericanconsumerswhoare only familiar with Priceline because of its catchy “Priceline Negotiator”commercials starring the actor William Shatner probably have no idea thatnearlytwo-thirdsofPriceline’srevenuescomefromhotelbookingsoutsidetheUnitedStates.In2015,Pricelineactuallyhadthehighestten-yearreturnofanystockintheFortune500.Aswithdirectblitzscaling,succeedingwithanM&Astrategyrequireshaving
a rareorunique insight into themarket;hadall theplayers in the travel spaceknownthevalueofonlinehotelbooking,Pricelinewouldn’thavebeenable toaffordtheBooking.comacquisition.Establishedplayerscanalsouseastringofacquisitions to blitzscale, provided they do a better job of integration thanHomeAway did. For example, Facebook’s acquisitions of Instagram andWhatsApp helped it fend off a dangerous competitor in Snap to achieve adominantpositioninsocialnetworkingforayoungergeneration.But established companies also have a number of disadvantages—beyond
simplybeinglessquickandnimble—totakeintoaccountwhenimplementingablitzscalingstrategy.
DISADVANTAGE#1:INCENTIVES
Amajor issue facedby establishedplayers is that the incentives tend to favorcautious expansion rather than aggressive blitzscaling. Successful companiesgenerallyassumethat theyalreadyhavesomethingvaluable,whichmeansrisktakingtendstobepenalized.Ifyoumaketheplayandfail,you’vedestroyedavaluable thing. That’s not something a start-up faces—a start-up is dead bydefault, so there is nothing to lose. Companies also face pressures fromshareholders,analysts,andthepress.Thethingis,bigcompanyleadersarenotwrongtobecautious!Alargeandpublicfailurecantankanestablishedplayer’sstockprice—anditsreputation.Moreover,thepotentialrewardshavetobehugetomatter.A$10millionopportunitythatastart-upmightseeasalife-changingbetispocketchangetoabigcompany.The incentives that drive individual employees can also have a negative
impactonblitzscalingattemptsinsideanestablishedcompany.Theemployeeorexecutivewhoproposes a riskyblitzscaling initiative is theonewho stands to
gain the most (promotions, bonuses, clout, etc.) from its success. In contrast,other employeesgain little from that success, andmight evenendup losing ifthatsuccessallowsitschampiontojumpoverthemforpromotionsorbonuses.And if the initiative is unsuccessful and costs the company a large sum ofmoney,itsemployeesallbearthecostoffailureaswell.Isitanywonderthatsomanyboldinitiativesarekilledincommittee?
DISADVANTAGE#2:UNSTAGEDCOMMITMENT
Another(largelyself-inflicted)disadvantageforlargecompaniesistheinabilityorunwillingnesstostagetheirinvestments.Thisresultsfrominternalincentives,which tend torewardmanagersbasedon therevenues that theyoversee,whilepenalizingfailureandundervaluinggrowthopportunities.Stagedinvestmentsletmanagersminimize theircompanies’downsideswhenperformingexperiments.Butsincemostsuchexperiments fail,bigcompanymanagers try to reduce therisk of failure by committingmore resources.Unfortunately,most blitzscalingopportunitiesaresoriskyanduncertain,andrequiresomuchcapitaltosucceed,thatmaking an unstaged commitment effectively bets the company.Thatmaynotbeabigdealtoastart-up,whichmustsucceedordie,butitisaverybigdealfor agoingconcernwhosebusinessmightotherwisechugalongprofitably foryearsorevendecades!Seniormanagersmayalso,foregoreasons,prefersplashyannouncementsand
major commitments to small experiments that can be blitzscaled if successful.Larger companies also impose a great dealmoremanagerial overhead; by thetimeyougetapprovalonaproposal,competitorsmighthavealreadylockedupthemarket.It should come as no surprise that one big company that continues to value
and encourage staged commitments and experiments is Amazon. Jeff Bezostalkedabout this inoneofhis famedshareholder letters.Bezoswants tomakesure that Amazon continues to have the start-up mindset, which he calls“Day1.”Bezoswrites,“StayinginDay1requiresyoutoexperimentpatiently,accept failures, plant seeds, protect saplings, and double down when you seecustomerdelight.”
DISADVANTAGE#3:PUBLICMARKETPRESSURE
Finally,establishedcompaniesthatarepubliclytradedfaceanadditionalsetofpressures to deliver short-term (i.e., quarterly) financial results. Blitzscalinggenerallyrequiressacrificingshort-termefficiency(andthusfinancialresults)toachieve long-termvalue creation.Privatelyheld companies areusually closelyheld; thiscanmake it easier toget themajor shareholders toagreeona risky,long-terminvestment—ifyouhaveshareholderswhoarewillingtoincurriskforachanceatamuchgreaterreward.Butawidelyheldpubliccompanymayfaceactivistinvestorsandothersuchshareholderrebellionsifitattemptstocarryouta blitzscaling strategy. This could even lead to the worst of possibilities—incurring the initial expense of blitzscalingwithout the necessary commitmentand follow-through to reap the long-term rewards. Many publicly tradedblitzscalerslikeGoogleandFacebookhavetriedtoavoidpublicmarketpressurebyissuingtwoclassesofstocksothatdecision-makingauthorityisvestedinasmallnumberofpeople(i.e.,Larry,Sergey,andMark).With these advantages and disadvantages in mind, here are a few specific
managementtechniquesor“hacks”thatlargecompaniescanusewhentheysetouttoblitzscale.
BLITZSCALINGHACKS
Oneproductivehacktohelpyourexistingcompanyblitzscaleistofindwaystoleverage people and businesses with prior blitzscaling experience. Oneobvious play is to partner with a blitzscaling start-up. For example, GMresponded to the riseofUber and the corresponding threat it represents to themarket for cars for human drivers by investing $500 million in Lyft, Uber’sblitzscaling rival. GM also hedged its bets by acquiring Cruise for its self-drivingcartechnology.A less obvious technique is to leverage the knowledge of venture
capitalists. Venture capitalists are keen fans of blitzscaling and the returns itbrings,eveniftheydidn’tknowthespecifictermbeforethebookcameout.Ifyouaskthemtobecomeminorityinvestorsinyourproject,theywillprovidearealistic assessment of your situation. For example, many large companiesmisprice their own assets or overvalue their own advantages, and attempt toblitzscaleeven if anobjectiveobserverwouldconsider theattempt ill-advised.Approaching venture capitalists is a quick way to get a sense of howknowledgeableprofessionalsassessthevalueofyourassets.
Onefinalway tomitigate the inherentpitfallsofblitzscalingwithina largerorganization is to treat thenew initiative as a companywithin a company.Once your blitzscaling project is under way, you will need to manage itdifferently than your regular projects. Blitzscaling’s increased pace anddecreased efficiency can seem reckless and wasteful when evaluated againstconventionalinitiativesthataredesignedtoprovidesteadygrowth.Asaresult,ablitzscalingprojectneedstobeinsulatedfromtherestofthecompanysothattheexecutive in charge can run it effectively.The classic example isSteve Jobs’sapproach tomanaging theoriginalMacintosh team,whichhadseparateofficesthat were off-limits to regular Apple employees. More recently, Larry PageappliedthissametechniquetoAndroidbyallowingAndyRubin’steamtoworkin separate offices—Google employee badges didn’t grant access to Androidoffices—andadoptdifferenthiringpracticesfromthoseoftheparentcompany.MuchthesamewastrueforthePlayStationprojectatSony,theKindleprojectatAmazon,andtheWatsonteamatIBM.
BLITZSCALINGBEYONDBUSINESS
Whilewe’vefocusedontheapplicationofblitzscalingintheworldofbusiness,thebasicprincipleofsacrificingefficiencyforspeed in thefaceofuncertaintycanbeappliedinjustaboutanycontext.Let’s consider how the growth factors and growth limiters of blitzscaling
translatetononbusinesssettings.
MARKETSIZE
Inthenonprofitworld,weneedtofindnewmeasuresofmarketsize,sincewecan’t rely on financial metrics like revenue. Often, the best measure mightsimplybe thenumberofpeoplewhose livesare improved,butothermeasuressuchas“yearsofhealthylife”or“metrictonsofcarbonsequestered”couldalsoservethisrole.Whilethemetricsmightchange,theprincipleofmarketsizestillapplies–without a large market, it makes little sense to blitzscale. One of themainreasonsthattheBill&MelindaGatesFoundationdecidedtotacklemalariapreventionandtreatmentistheenormoussizeofthemalaria“market.”In2012,207 million people suffered from the disease and 627,000 died from it, with77percentofthosedeathsaffectingchildrenundertheageoffive.Thosefigures
include a 42 percent reduction in annual deaths from2000 to 2012, thanks inpart to the efforts of the Gates Foundation. That is a largemarket where theabilitytoblitzscalemakesahugeimpact.
DISTRIBUTION
Distribution is justascriticaloutsideof thebusinesscontextas it is forprofit-seeking companies. No matter how potentially effective your “product”—whetherthatproductisasocialservice,apoliticalcandidate,oranythingelse—isatimprovingthelivesofthosewhoadoptit,itsimpactisdirectlyproportionalto your ability to execute an effective distribution strategy. The MozillaFoundationwasn’t theonlyopen-sourceorganization to create aWebbrowser(Firefox), but itwas the only nonprofit organization thatwas able to leveragedistributiontoachievealeadingmarketshare.In2008,BarackObamawonthepresidencyinpartbecausehiscampaignwasthefirsttoleveragethedistributionpossibilities of the Internet, including leveraging existing grassroots networksandachievingviralityviasocialmedia.
GROSSMARGIN
Sincemanynonprofitsdon’tchargethepeopletheyserve,grossmargindoesn’tapply.However,wecanusemetrics that alignwith the spiritofgrossmargin,suchaseconomicimpact.Atahighlevel,grossmarginisameasureofimpactperdollar, and thegreater the impactperdollar, themoreamenableanot-for-profit business should be to blitzscaling. For example, the International CivilSociety Support estimates that each $1 spent on malaria prevention andtreatment generates $20 of economic benefit, with insecticide-impregnatedmosquito bed nets representing the single most cost-effective intervention.That’s the kind of impact that compares favorably even to software grossmargins.
NETWORKEFFECTS
Networkeffectsarerelativelyrareinthenonprofitworld.Whiletherearemega-NGOs such as the Red Cross and the United Way, their market position is
largelyduetoeconomiesofscaleratherthantruenetworkeffects.Butitisstillworthconsideringwhetherornotitispossibletotapintonetworkeffects,sincedoingsocanhavesuchamajorimpact.Forexample,SalKhan’sKhanAcademybeganwhenSalstartedtutoringone
ofhisyoungcousinsovertheInternet.Whenothercousinsstartedsigningup,hedecidedtoposthis lecturesonYouTubesothatanyoneintheworldcouldusethem.Thecriticaldecision to leverage theYouTubeplatformmeant thatKhanAcademy had both an enormousmarket (anyonewho could accessYouTube,whichistosay,mostofhumanity)andapowerfuldistributionplatform(anyonesearching for educational content on YouTube was likely to run across KhanAcademy). As the Khan Academy gained a massive user base, it began tobenefitfrombothindirectandstandard-basednetworkeffects.EducatorsbeganincorporatingKhanAcademyvideosintotheirofficialcurriculum,andcreatinglessonplansthattheysharedwithothereducators.Today,theKhanAcademyisused by 40 million students and 2 million educators every month (the entireUnited States has only 50.7 million K–12 students), and volunteers havetranslateditsvideosintothirty-sixlanguages.
LACKOFPRODUCT/MARKETFIT
Inthecaseoffor-profitbusinesses,theremorselesslogicofthemarketeconomyquicklyeliminatescompaniesthatfailtoachieveproduct/marketfit.Withouttheabilitytoachievetraction,businesseslacktherevenuetosurviveandhavelittleability to raise additional funding from investors. In contrast, nonprofits oftenreceive grants and donations for noneconomic reasons, and the flow of fundsisn’talwayscorrelatedwith theeffectivenessof theorganizationbeingfunded.The“clients”are thepeople itserves,but the“customers”are thefunders.Yetproduct/market fit is still important for those organizations if they want toblitzscale. In general, the more effectively a nonprofit organization serves its“clients,” the better it is able to raisemoremoney from“customers.”Charity:Water is a nonprofit that provides clean, safe drinking water to people indeveloping nations. It is also amodel of product/market fit for its users,whohavebenefited from themore than twenty-three thousandwaterprojects it hasfunded,andforitsfunders,whocanseephotosofthewellsitbuildsandknowthat 100 percent of their donations will go to fund those projects (theorganization’s operating costs are covered by foundations and sponsors).As a
result, inthedecadesinceitsfoundingin2006,Charity:Waterhasraisedover$252millionfrommorethanthreehundredthousandindividualdonors.
OPERATIONALSCALABILITY
If anything, operational scalability is a bigger challenge outside the businessworld.Asabusinessscales,itusuallyhastherevenuesorVCdollarstoinvestheavilyinascalableinfrastructureortohireadditionalemployees.Thebusinessworldisalsofullofcompaniesthathavesuccessfullyscaled,whichmeansthatitis easier to bring in employees who can help you manage rapid growth. Incontrast,nonprofitsusuallydon’thave thesamefinancialcapital,andcertainlydon’thaveaccesstothesamekindofexperiencedhumancapital.Thisplacesagreaterpremiumondesigningabusinessmodel thatdoesnot require asmanyresourcestoscale,asinthecaseofanopen-sourceorganizationliketheMozillaFoundation.Asidefromthegrowthfactorsandgrowthlimiters,oneothermajorpotential
differenceishownonprofitsandimpactorganizationslookatcompetition.Inthebusinessworld(andcertainnonbusinessorganizationslikepoliticalcampaigns),competition—specifically, beating that competition—is one of the mostimportantmotivationsforblitzscaling.Incontrast, theGatesFoundationwouldwelcomeanothermajorplayer thatspentbillionsofdollars to try to“beat” theGates Foundation to achieving the goal ofmalaria eradication. Thismay helpexplainwhyblitzscalinghasbeen relatively lesscommon in theworldbeyondbusiness. However, given the scale of the challenges before us today, fromclimatechangetopovertytoaneducationsysteminneedofreform,webelievethat the time is ripe to consider how to apply scalable technology solutions totraditionallyunscalableissues.Let’s examine twovery specificbutverydifferent examples to see someof
thewaysinwhichyoucanapplytheprinciplesofblitzscalingbeyondbusiness.
Example#1:DressforSuccess
DressforSuccess(DFS)helps low-incomewomenget jobsbyproviding themwith donated professional attire and coaching them through the interviewprocess. Ninety-nine percent of its operating funds come from grants,
governmentfunding,anddonations.Without investordollarsorrevenues tofundexpansion,DFSfounderNancy
Lublin had to find clever ways to circumvent the challenges of operationalscalability without consuming cash. One particularly clever strategy was toleverageitsinfrastructurelimitationstogetaroundthelimitationsinmanpower.Toachievescale,DFSneededtofindawaytoscreenpotentialclients(toensurethe organizationwas serving thosewhomost needed it) and staff its clothing“shop,” both of which would normally require either paid employees orextensive recruitingofvolunteers. Instead,Lublinpartneredwithorganizationslike domestic violence shelters that were serving the same clients, andaccomplishedbothatzerocost.DFSonlyacceptedclientswhowerereferredbypartners;inexchange,thosepartnerswererequiredtoprovidevolunteerstohelpstafftheshops.ThisallowedDFStoscalethenumberofpeopleitservedalongwiththeworkforceneededtoprovidethatservice,allwithoutcostingitadime!Lublin also leveraged the power of an innovative distribution model by
“franchising”DFS.AnyonewhowantedtostartaDFSshopwasinvitedtoflytoNew York and stay on a futon in her apartment. She then trained thoseentrepreneursandsentthembacktotheirhometownstostartupnewDFSshops.WhenLublinleftDFSin2002,shehadexpandedtheorganizationtoseventy-sixshops.Today,DFShasexpandedintotwenty-onedifferentcountriesandhelpednearlyonemillionwomen.
Example#2:BarackObamaforPresident
In 2008, the presidential campaign of Barack Obama used the power ofblitzscaling—especially business model innovation—and the tools of SiliconValley to catapult a little-known first-term senator from Illinois to theWhiteHouse, despite running against a series of prominent national politicians,includingformerFirstLadyandfellowsenatorHillaryClinton.The key business model innovation behind Barack Obama’s long-shot
campaignforthepresidencywasanunprecedenteduseofconnectivitytoenableandcoordinateadecentralizedmovement.ObamaannouncedhiscandidacyonFebruary10,2007.AccordingtocampaignadviserSteveSpinner,thecampaigngrew from zero to seven hundred employees (and orders of magnitude morevolunteers) in a single year. The key to this rapid growth was the use oftechnology to leverage large existing networks and achieve powerful
distribution.First, Obama focused on raising small donations from individuals via the
Internet rather than raising large donations from well-heeled, traditionalDemocratic Party donors. This was partially out of necessity, since he wasrunning for theDemocratic nomination againstHillaryClinton,whowas bothviewed as a presumptive favorite and had deep relationships with big donorsfromheryearsintheWhiteHouseandtheSenate.Butasitturnedout,thisnewbusiness model allowed Obama to raise more money than any previouscandidate: more than $650 million in campaign contributions, nearly$300millionmore than the previous record,which had been set by PresidentGeorgeW.Bush during his 2004 reelection campaign.Over half that amountcamefromdonationsoflessthan$200;incontrast,only27percentofthemoneyraisedduringthe2004electioncyclecamefromsmalldonors.Second,theObamacampaignalsoutilizedtechnologytobuildandmanagean
army of volunteers to get out the vote. Here, the Obama campaign benefitedfroma remarkablepieceof luck; shortlybeforeannouncinghis candidacy, thecampaignreachedouttotheyoungsocialnetworkFacebooktosetupanofficialpage. The person they reached, Facebook cofounder Chris Hughes, becameconvinced thatBarackObamacouldwin thepresidencyandchange theworld,andHughes leftFacebook to join the campaign.Hughesbroughtwithhimhispersonalexperiencesworkingatoneoftheworld’sgreatblitzscalingcompanies,andmovedquicklytoapplythetoolsofSiliconValleytotheObamacampaign.Hughesandhisteamendedupcreatingthreekeytoolsthatleveragedgrowth
factorstohelpObamawintheelection.Thefirstwasmy.barackobama.com,orMyBOforshort.MyBOwasasocialnetworkthat leveragedexistingnetworksof Obama supporters, allowing them to connect with one another, as well ascreate groups, plan events, and raise funds.Over the course of the campaign,volunteersusedMyBOtocreatetwomillionprofiles,hosttwohundredthousandoff-lineevents,andraise$30million.ThesecondtoolwastheNeighbor-to-Neighborcanvassingtool.WhenMyBO
usersloggedin,Neighbor-to-Neighborprovidedalistofundecidedvoterstheycouldcallorvisit.Neighbor-to-Neighbortappedintoonlinedatabasestomatchvolunteers with people they would likely connect with, taking into accountfactors like age, profession, languages spoken, andmilitary service.Neighbor-to-Neighbor generated some eight million calls—and tremendous word ofmouth.
The final tool was the Vote for Change voter registration site, whichautomatically sorted out the fiendishly complicated nest of local voterregistrationrulestohelppotentialObamavotersregistercorrectly.Forexample,collegestudentswouldloginandbeaskedforthelocationofboththeircollegeandchildhoodhome;VoteforChangewouldthenhelpthemregisterinthestatewhereastudent’svotewasmorebadlyneeded.Duringthecampaign,VoteforChange helped onemillion people register to vote—roughly the same numberthat two thousandpaidstaffcouldhandleusing theold-fashioneddoor-to-doormethod.OnTuesday,November4,2008,BarackObamawaselectedtheforty-fourth
president of the United States of America. Thanks in part to the blitzscalingtechniques his campaign employed, he received over sixty-ninemillion votes,stillarecordforanyUSpresidentialcandidate.As these examples demonstrate, blitzscaling can be as powerful a tool for
social impact and change as it can be for building a massively profitablebusiness.It’snoteasy;you’llneedeitheraccesstosignificantcapital(aswesawwithBarackObama’s 2008 presidential campaign), the ability to leverage thecontributionsofacommunityoranotherexistingnetwork(aswesawwithDressforSuccess),orboth.Butifyouareabletosupportthiskindofrapidgrowth,thelessons of blitzscaling can help you manage the strains of that growth andmaximizetheimpactyouhaveontheworld.
BLITZSCALINGINGREATERSILICONVALLEY
Oneoftheinterestingdevelopmentsinthebusinessworldoverthepastdecadeishowotherhigh-techecosystemsonthePacificcoastoftheUnitedStateshavebecomemore tightly integratedwithSiliconValley.Formost of the twentiethcentury, Seattle, Los Angeles, and Silicon Valley were very different anddifferentiated industry hubs. While Silicon Valley specialized in computers,SeattleandLosAngelesbothhadstrongaerospaceanddefense industries,andeach ecosystem also boasted its own market leadership positions in coffee(Seattle) and the entertainment industry (LosAngeles).But in the twenty-firstcentury, Seattle and Los Angeles have also become home to high-techecosystemsthatareincreasinglytiedtoSiliconValley.Ina2017articletitled“HowAmerica’sTwoTechHubsAreConverging,”the
Economist argued that Seattle and SiliconValleywere becoming increasingly
intertwined, citing as evidence the fact that most of the venture capitalinvestmentinSeattlestart-upshascomefromSiliconValleyVCfirms,andthataboutthirtySiliconValleyfirmshadopenedSeattleofficestotapintothatcity’splentiful supply of computer scientists, while Seattle’s two dominantblitzscalers, Amazon andMicrosoft, have thousands of employeesworking inSilicon Valley. Meanwhile, Seattle’s Amazon Web Services has become thecloudcomputingplatformofchoiceforSiliconValleystart-upsandscale-ups.LosAngeleshasalsoseenimpressivegrowthasastart-upandscale-uphub.
AccordingtotheresearchfirmCBInsights,LosAngeles–basedstart-upstookin$3 billion in funding in 2016, up sixfold since 2012. The LA ecosystem,including the so-called Silicon Beach area along the Pacific coastline, hasproducedanumberofimportantcompanies,includingSnapandSpaceX,eachofwhich is worth over $10 billion, as well as other success stories like DollarShave Club. It is worth noting that Snap’s founders met while studying atStanford, thatSpaceXwas foundedby formerSanFranciscanElonMusk,andthatthemajorityoftheirinvestorswereSiliconValleyventurecapitalfirmslikeLightspeed, Founders Fund, and Venrock. Like Seattle, Los Angeles is anoperations base for major Silicon Valley companies such as Google. Theserelationships aren’t particularly surprisingwhenyou consider that bothSeattleandLosAngelesarecloseenoughtoSiliconValleytoallowtheintertwiningoftheirnetworksofcapital,talent,andlearning:eitherashortflight—Seattleistwohoursaway,LosAngelesjustanhour—orasix-hourdrivefromLA,withplentyof Tesla Superchargers along I-5 so that venture capitalists can make thejourney!Thishelpstointegratethecapitalnetworks;it’seasyforSiliconValleyinvestors to invest in Seattle-and Los Angeles–based deals because they cansimply fly in and out for boardmeetings. This also helps integrate the talentnetworks, since entrepreneurs can easily travel between hubs to develop andmaintain relationships and to share insights and learning face-to-face. Forexample, geography allowsElonMusk to run bothTesla (SiliconValley) andSpaceX(LosAngeles).SeattleandLosAngelesalsooffergoodquality-of-lifebenefits to individual professionals, since they are major cultural centers andpopular tourist destinations, with less expensive housingmarkets than SiliconValley(thoughhardlyinexpensive).Thesetieswilllikelygetevencloserifthecitiesareconnectedviaadditional
transportation links likehigh-speed railorElon’sproposedHyperloop,orwiththeadventofself-drivingcars,allofwhichwouldmaketravelandcommutingbetweenthesecitiesandSiliconValleyfasterandcheaper.Thus,LAandSeattle
arebecomingincreasinglyfertilegroundsforentrepreneurship,andgoodplacestosetupacompanythatyouareplanningonblitzscaling.ItwillbeinterestingtoseeifAmazon’sHQ2projecttosetupasecondheadquarters(Amazonintendstospend$5billiontoconstructanewcorporatecampusthatwillaccommodatefifty thousand employees) ends up expanding “Greater Silicon Valley” evenfurther(andfarther).Moody’spickedAustinasthemostlikelycitytobechosen,whiletheNewYorkTimesbelievedDenvertohaveagoodchance.Eitherwouldbealogicalexpansionpoint,giventhelong-standingpatternofmigrationfromSiliconValleytoColorado,andthe“nerdbird”flights thatSouthwestoperatesbetweenSanJoseandAustin.
OTHERBLITZSCALINGREGIONSTOWATCH
Within the United States, cities like Boston and Austin have emerged aspowerful tech hubs, followed more recently by Boulder, Colorado, and evenNew York City. Over in Europe, cities like London, Stockholm, and Berlin(where the Samwer brothers’ Rocket Internet is attempting to industrializeblitzscalingasabusinessmodel,withnotablesuccessesandfailures)havebegunproducing interesting companies as well. According to research at Wharton,Stockholm actually produces the second highest number of billion-dollar“unicorn”start-upsafterSiliconValleyitself.Around65percentofworking-age(eighteen to sixty-four years old) Swedish adults think there are goodopportunities to start a company in Sweden, compared with 47 percent ofworking-ageAmericans.For example, Stockholm’s streaming music giant, Spotify, has a record of
blitzscalingthatmostSiliconValleyunicornswouldenvy.Spotify’scofounders,Daniel Ek and Martin Lorentzon, are both serial entrepreneurs with priorblitzscaling experience—Ek was the CTO of Stardoll, while Lorentzoncofounded Tradedoubler. Spotify uses the proven freemium business model,offering a basic free service and encouraging users to subscribe for higher-quality audio and zero ads. Since its launch in 2008, Spotify has pursued acourseofextremelyaggressive investment, raisingover$2.5billion infundingfromSiliconValleyVCs, such as Founders Fund,Accel, andKleiner PerkinsCaufield&Byers,aswellasglobalinvestorswithscaleexperience,suchasLiKa-shing’sHorizonsandYuriMilner’sDigitalSkyTechnologies(DST),whilegrowing from onemillion paying subscribers in 2011 to sixtymillion payingsubscribersin2017.
It’s worth noting that despite their success in Stockholm, in 2016 Ek andLorentzon began to fear that policies like restrictive and expensive housingregulations for immigrants and heavy taxation of stock options couldmake itdifficult for Spotify to remain there. Sure enough, in February 2017, Spotifyannouncedthatitwouldadd1,000newjobsinitsNewYorkoffice,makingtheUnitedStatesthehometothemajorityofSpotify’sworkforce.Outside theWesternworld, theprospectsmightbe evenbrighter.Weknow
aboutChina,ofcourse,butIndiaisalsoprojectedtoovertaketheUSeconomythis century. Indian e-commerce giant Flipkart has raised nearly $7.3 billionfrominvestorsaroundtheworld,includingAccel(SiliconValley),TigerGlobal(NewYork),Naspers(SouthAfrica),GIC(Singapore),andSoftBank(Japan).Itsfounders, Sachin Bansal and Binny Bansal—the two are not related—bothworked for Amazon. Africa is pioneering mobile services like the M-Pesamobilepaymentsystem,whichwasdeveloped in theUK,managedbyIBMinAmerica,andwhose technologyisnowmanagedbyHuawei inChina.M-Pesaaccounted for $28 billion in transactions in Kenya in 2015; for comparison,Kenya’s GDP that year was $63 billion. Latin America represents a fast-growing, largely Spanish-languagemarket. Israel,which boastsmore start-upspercapitathananyothernation,isaleadingcenterforcybersecuritycompaniesand is also home to a thriving venture capital community. EvenAustralia hasproducedsuccessfultechnologycompanieslikeAtlassian.Blitzscaling in emerging ecosystems poses both different challenges and
different opportunities. Emerging ecosystems lackmany of the platforms thatestablished ecosystems like Silicon Valley or, more broadly, the US marketprovide—for example, payment systems and shipping vendors, let aloneprofessional service providers (lawyers, accountants, etc.), experiencedexecutives, and aggressive venture capitalists. This makes blitzscaling moredifficultandleadstoslowerratesofgrowth.It’sfareasier to leverageexistingplatformsthantobuildyourown.Ontheotherhand,onceyouaresuccessful,havingbuiltyourownplatforms
represents a major competitive advantage, which often compounds over time,resultinginfasterlong-termgrowth.MercadoLibregrewmuchmoreslowlythanAmazonduringitsearlyyears.InLatinAmerica,fewerthanhalfofconsumerseven have a bank account. The company couldn’t simply tap into ubiquitouscredit card networks and established shipping vendors the way Amazon did;instead,ithadtobuilditsownpaymentandlogisticssystems.
Today, however, owning platforms like Mercado Pago, the leading e-commercepaymentsysteminLatinAmerica,allowsMercadoLibretosustainahigher growth rate while creating barriers to potential competition. WhereasrivalswhowanttocompetewithAmazonintheUSmarketcanstartandgrowquickly thanks to Visa and UPS, a rival toMercadoLibre would have to useMercadoLibre’spaymentandlogisticsplatforms,makingitmuchmoredifficulttogainsignificanttraction.MercadoLibrewasalsoabletotakeadvantageoflessonslearnedbyprevious
blitzscalerslikeeBay.In2001,eBayacquiredaFrenchcompanycallediBazar,whichhadaBraziliansubsidiary.BecauseeBaywanted to focusonEurope, itmadeMercadoLibreanoffer:takeovertheBrazilianoperationinexchangefor19.9percentofMercadoLibre.Thedeal includedbotha five-yearnoncompeteagreement (whichmeant thatMercadoLibre didn’t have to worry about eBayexpanding into theLatinAmericanmarket for at least that time period) and abestpracticessharingagreement.WhileCEOMarcosGalperindidthedealforthe noncompete agreement, during an interview for Reid’sMasters of Scalepodcast,he said that themost importantelementof thedeal forMercadoLibreendedupbeingthebestpracticessharingagreement:
What ended up being incredibly valuable for us was that really,reallyintensebestpracticessharingprocess.WewerebasicallylikeaneBaysubsidiaryforfiveyears!Wewouldgoupthere[toeBayHQ inSiliconValley]everyquarter, andeverydifferent sectorofourcompanywouldexchangebestpracticeswithdifferent sectorsofeBay.IthelpedusscaleandseeallthedifferentproblemseBaywas having in different parts of the world and with differentcompetitors.Wewere able to pick and choose; therewere thingseBaywas doing that we really liked, and things eBaywas doingthatwedidn’tthinkappliedtoLatinAmerica,whichwewoulddodifferently.
Marcosandhisteamdidn’tjustimitateeBay.TheylearnedfromeBay’sbestpractices,andadaptedthemtothespecificcharacteristicsoftheirownmarket.All these new ecosystems around the globe represent interesting and
potentially differentiated opportunities, much like China fifteen years ago orSiliconValley twenty-five years ago.Bostonhaswon a leadership position in
health,forexample,becauseofitsworld-classhospitalsanduniversities,whileNewYorkistheleaderinfashion-relatedbusinesseslikeRenttheRunwayandBirchbox.Countries likeEstonia havemade their dependence on internationalmarketsastrength;Skype(foundedbyEstonianprogrammersPriitKasesaluandJaan Tallinn) wasn’t likely to have started in the United States becauseinternationalphonecallssimplyweren’tasimportanttotheUSconsumer.
CHINA:THELANDOFBLITZSCALING
RememberPonyMa’sdecision to launchandthenblitzscaleWeChat in2010?HedidthisdespitethemassiverisksthatWeChatwouldposetohismatureQQdesktopproduct,andthustoTencent’soverallbottomline.Bytakingtheserisksand launchingWeChat,Mawasable to rejuvenatehiscompanyandpush it toevengreaterheights.ThestoryofWeChatexemplifieswhy,ifanything,Chinamayverywellend
up being an even better ecosystem for blitzscaling than Silicon Valley. LikeSiliconValley,Chinahasanentrepreneurialculturethatencouragesrisktaking,ahighlydevelopedfinancialsectorthatiswillingtofundaggressivegrowth,anda plentiful supply of high-value technology talent.But thanks to its incrediblerecentgrowth,China’smarketisbothmassiveandopentodisruption.Chinahasbeenoneoftheworld’sfastest-growingeconomiesfordecades,and
PricewaterhouseCoopersprojectsthatChina’seconomywillovertakethatoftheUnitedStatesinsizeby2030.Inmanyareas,italreadyhas.In2016,thevolumeofmobilepayments inChinawas$8.6 trillion. Incomparison, thesamefigurefortheUnitedStateswas$112billion.Inotherwords,China’smobilepaymentsmarketwasnearlyseventy-seventimesthatoftheUnitedStates.DidiChuxingprovides twentymillion ridesperday inChina,over triple thevolumeofUberworldwide.ThesefactorsgiveChinaamajoradvantageoveralmosteveryotherecosystemwhenitcomestothegrowthfactorofmarketsize.China also hasmajor advantages when it comes to overcoming the growth
limitationsofoperational scalability, thanks to its flexible labormarket,whichwas explored in a 2012New York Times article aboutApple’smanufacturingoperations in China: “Apple’s executives had estimated that about 8,700industrialengineerswereneededtooverseeandguidethe200,000assembly-lineworkerseventuallyinvolvedinmanufacturingiPhones.Thecompany’sanalystshad forecast itwould take as long as ninemonths to find thatmany qualified
engineersintheUnitedStates.InChina,ittook15days.”The result is an ecosystem in which companies grow, break apart, and
recombinewith incredible speed. “Innovationmoves faster here,” saidKai-FuLee,whorunsSinovationVentures,andusedtobetheheadofGoogle’sChinaoperations.TheChinesemarketviewsgrowthasthefirst,last,andbestsolutiontoalmostanyissue.PerhapsthisiswhyChinesestart-upstendtoscaleupatanevenfastertempothanSiliconValleyfirms.For example, it took less than five years for Chinese smartphone maker
Xiaomi togo from founding to theworld’smostvaluable start-up in2014—ithas since been surpassed byUber andDidi Chuxing, no blitzscaling slouchesthemselves.Lei Jun foundedXiaomi in2010;by2015, itwas the third-largestsmartphonemanufacturerintheworldafterSamsungandApple.ButjustascompaniescanrisefasterinChina,companiescanalsofallfaster
in China. In 2016, IDC reported that Xiaomi’s sales dropped 40 percent yearoveryear, as itsonline-only sales strategybegan to falterandcompetitors likeOPPO and Vivo gained market share by selling through brick-and-mortardistributors.AtleastoneanalystpredictedthatXiaomi’svaluewoulddropover90percent.Xiaomi’sresponsetothiscrisisdemonstratesboththefiercecompetitiveness
ofLeiJunandtheamazingtempopossibleinChina.Thecompanyattackeditsdistributionproblemswith a rapid,massive effort to buildup its off-line saleschannel, opening one hundredMi Home retail stores in a single year, with atargetofopening two thousandstoresby2019. In the firstquarterof2017,anastounding34percentofXiaomi’ssmartphonesalesinChinacamefromitsonehundredretailstores,which thecompanyclaimsgeneratesalespersquarefootsecond only to Apple’s famed Apple Stores. In 2017, IDC reported thatXiaomi’ssaleshadrebounded59percentfromthepreviousyear,placingitbackamongtheworld’stopfivesmartphonemakers.That’saragstorichestoragstorichesstory,allcompressedintolessthanadecade.Asfounders, investors,andauthors,wehaveapersonal, in-depthfamiliarity
withtheSiliconValleyway;incontrast,ourknowledgeofChinaisnecessarilythatofoutsiders.Yetwecan’thelpbutbestruckbyhowmanyvaluablelessonsthesetwoecosystemscanlearnfromeachother.Forexample,China’sspeeddemonstratesthevalueofintensecompetitionas
a motivator. On one occasion, Xiaomi’s Lei Jun told me, “You Americanentrepreneursarelazy.Thevastmajorityofmycompanyisstillworkingatnine
o’clock on a Saturday night.” In some ways, he’s right. Chinese blitzscalersworkwithanintensitythatfewinSiliconValleycanmatch.RatherthanstayingopenduringthestandardAmericanbusinesshoursof9a.m. to5p.m.,Xiaomioperatesona“996”model—getinat9a.m.,leavetheofficeat9p.m.,andworksix days a week. I saw the same thing at LinkedIn China. To make a tightdeadlineforour“RedHorse”project,ourChinateamleaderDerekShensimplymovedtheentiredevelopmentteamtoahotelfortwoweekssothatitsmemberscouldworkaroundtheclockwithoutanyofthedistractionsofnormallife.A by-product of this intense work ethic is that it allows for much faster
decisionmaking, a key advantage in blitzscaling.When I interviewedhim formy Masters of Scale podcast, Andrew Ng, a professor at Stanford whocofoundedCourseraandledmajormachinelearningeffortsatGoogleandBaidu(China’sleadingsearchengine),toldmethatwhenhewasatBaidu,heoncehadanHRquestioncomeupwhilehewasatdinner.HetextedhisHRleadat7p.m.;shetextedherteamforinput,andby7:30,Andrewhadhisanswer.“Ifshehadtaken longer than an hour to respond,” Andrew said, “I would have gottenworried.” That kind of rapid decision making might make many feeluncomfortable, but by consistently making quick decisions, Chineseentrepreneursbecomecomfortablewiththediscomfortanduncertainty,allowingthemtomoveevenfaster.Another advantage comes from China’s massive talent pool. The sheer
abundance of human capital allows companies in China to scale theirorganizationsmorequickly,includingopeningmultipleofficesinmultiplecities.China alsohas a thingor two to teachSiliconValley about tapping the entiretalentpool.Forexample,Chinahasprovenanamazingenvironmentforwomenentrepreneurs. Of the seventy-three women in the world who are self-madebillionaires, forty-nine (over two-thirds!) live in China. Eight of the tenwealthiestself-madewomenintheworldareChinese.Finally, China’s relatively recent rise into its status as an industrial power
meansthatmoreofitsindustriesarestillnascentandthusupforgrabs.WhereSiliconValleymighthavecorneredthemarketonsoftwareandtheInternet,withasidelineinhardware,Chinahasfast-growingcompaniesineveryindustryfromfarmingtochemicals.Yet despite these impressive strengths, China can also learn from Silicon
Valley.Forone thing,SiliconValley’scomparatively less freneticpacemeansthat it can pursue deeper tech and longer time horizons like Elon Musk’s
commitment to interplanetary travel and Google’s famous $750 millioninvestmentinCalico,aprojectto“curedeath.”SiliconValleystillhasaleadinmost deep technology innovations such as artificial intelligence (AI), virtualreality(VR),spaceflight,andnuclearpower.WhileSiliconValleyiscertainlyhometoplentyofruthlesscompetition, the
culture also encourages more collaboration between companies. Thiscollaboration leverages network connections across companies to drive greaterinnovation and productivity for the entire region.WhenGoogle open-sourcedtheTensorFlowsoftwarelibraryin2015,itallowedGoogletoleverageexternalbrainpower to improve its machine learning projects, but it also allowedcompanies throughout SiliconValley (and the rest of theworld) to acceleratetheirownmachinelearningprojectsaswell.Moreover,SiliconValley’shistoryofblitzscalingputitseveraldecadesahead
of China in terms of concentrated experience and institutional knowledge.Remember, half of the world’s most valuable technology companies areclusteredinthissmallregionwithapopulationoflessthanfourmillion.That’sabout10timessmallerthanthemetropolitanareaofGuangzhou,and350timessmallerthanthepopulationofChina.Meanwhile,China’stwocompaniesworthmore than $100 billion,Alibaba andTencent, are both less than twenty yearsold. Together, these facts mean that despite China’s massive labor pool andincrediblesupplyoftechnicaltalent,itlacksthedensityofSiliconValley,andisstill limitedintermsofthebenchstrengthofscaleexecutivesavailabletohelpmanageblitzscalingcompanies.Finally, China’s more insular management and hiring practices can present
obstaclestoblitzscaling.MyfriendJerryYang,whocofoundedYahoo!andledthatcompany’sprescientinvestmentinAlibaba—whenJerrymadehisfirsttripto China in 1997, the guide the Chinese government provided him was anEnglish teacher named JackMa—has observed that Chinese companies try tobreed their leaders fromwithin the organization. Unlike in Silicon Valley, inChina seniormanagers are rarelybrought in fromexternal companies, and thefewthathavebeenhiredtypicallyhaven’tworkedoutwell.Forexample,HugoBarrawas awell-regarded executive atGooglewho joinedXiaomi as its vicepresident of international, but left the company after a little over twoyears toleadFacebook’sVRefforts.This internal approach has major implications for blitzscaling; you have to
start thinking about how to fill leadership positions years in advance and start
grooming people for them right away. This also means that there is far lessmobility between firms, and, as a consequence, less intermixing of ideas andinnovation.Thismaybechanging;JerrypointedoutthatthefirstgenerationofChinesestart-upgiantsisalreadybeginningtoprovidetheseedsofthenext.Forexample, ChengWei, the founder of ride-hailing giant Didi Chuxing, learnedhowtoscaleatAlibaba,whereheworkedforeightyearsbeforestartinghisownfirm. That experience probably helpedCheng scaleDidi at a pace thatmakesUberenvious.DespitebeingfoundedthreeyearslaterthanUber,Didicompletedmore rides in 2015 than Uber had during its entire existence to that point.Meanwhile, the fact that first-generationcompanies likeAlibaba,Tencent, andBaidu are all Didi investors affords Didi’s management access to knowledgenetworkshelpfulforblitzscaling.Overall,wethinkChina’stechnologyindustryleadersaredoingagoodjobof
learningfromSiliconValley.WhenItravelandspeakinChina,Ifindthatmyaudiencesarefamiliarwithwhat ishappening inSiliconValley.MostChineseexecutivesspeakand readEnglish,andare reading the latestEnglish-languagenews on a daily basis. How many American or European executives readChinese and are staying abreast of developments inChina? If youwait for aninnovation tomake itsway into theEnglish-languagepress,perhapsbecauseaSilicon Valley company is now doing it, you might be giving China’sblitzscalersaone-yearheadstartontheglobalmarket.ThebiggestopportunityisforSiliconValleyandChinatoworktogetherand
combine their respective strengths. According to Andrew Ng, it took acombination of ideas from both sides of the Pacific to drive breakthroughprogress in speech recognition.SiliconValleycompanies likeNvidiaprovidedthegraphicalprocessorunits(GPUs)topowermachinelearningnetworks,whileprogresscamefromcombiningSiliconValley’sexpertiseinGPUprogrammingwithChina’s expertise in supercomputing.As ofNovember 2016, theworld’smost powerful supercomputer was the Sunway TaihuLight at the NationalSupercomputingCenter inWuxi,China,while number twowas the Tianhe-2.America’s most powerful supercomputer, the Titan at Oak Ridge NationalLaboratory in Tennessee, was less than one-fifth as powerful as the SunwayTaihuLight.There is no telling what kind of wealth and progress might emerge out of
futurecollaborationbetweentheleadinginnovatorsinthesetwoecosystems.
DEFENDINGAGAINSTBLITZSCALING
Thusfar,we’vefocusedonhelpingyouunderstandhowyoucanuseblitzscalingtobuildastart-upintoascale-up,ortorapidlyscaleanewproductorbusinessunit.Inotherwords,you’velearnedhowtouseblitzscalingtoplayoffense.Butwhatifyou’retheincumbent?Whatif,insteadofhavingnothingtolose,
andeverythingtogain,theoppositeistrue?If you find yourself in a positionwhere competitors are trying to blitzscale
yourexistingbusinessoutofexistence,youhave threebasicoptions todefendyourself:beatthem,jointhem,oravoidthem.
OPTION#1:BEATTHEM
The firstoption todefendagainstblitzscaling is tobeat thembycontinuing toplayyourtraditionalgame.Aswe’vediscussed,manyattemptstoblitzscalearedoomedtofailure.Youshouldassessthegrowthfactorsandgrowthlimitersofthebusinessmodel,andiftheyseemill-suitedtoblitzscaling,notoverreactingisprobablyyourbeststrategy.Fansof the lateMuhammadAlimay recall his “rope-a-dope” strategy from
his“RumbleintheJungle”boxingmatchagainstGeorgeForeman.Therope-a-dopecallsforallowinganopponenttopunchhimselfout;whenthatopponentisexhausted,youcanbeathimwithacounterattack.InthecaseofWebvanduringthedot-comboom,themanyproblemswithits
businessmodel(lowmargins,massiveoperationalscalabilityissues)meantthatitsattempttoblitzscalewasprobablydoomedfromthestart.Establishedgrocersessentially rope-a-doped: they set up their own online grocery initiatives, butthese were incremental, low-investment efforts. Safeway even exploitedWebvan’s failure by allowing Webvan to convince early adopters to ordergroceries online and then setting up its own grocery delivery service to servethosestrandedcustomers.Of course, those same grocers now face amuch different competitor in the
formofAmazonanditsacquisitionofWholeFoods.Thesecircumstancesseemtocallforaverydifferentresponse.Amazonisunlikelytopunchitselfout!
OPTION#2:JOINTHEM
Ifyourmarketdoesseemripeforblitzscaling,oneobviousresponseistolaunchyourownblitzscalingeffort.Theproblemwithdoingso,especiallyifyou’reanestablishedcompany, is thatyoumightnothavethetechnologyorexpertise towin a head-to-head competition. Youmight be able to buy the technology orexpertise,butthisbringsitsownsetofrisks.First, if blitzscaling is occurring, that almost certainlymeans that investors
(public or private) are enthused enough about the market to provide cheapcapital.Thatmeansthatanyacquisitionsarelikelytobeveryexpensive.Second, as I warned Brian Chesky about buying Wimdu, any merger or
acquisition bringswith it the possibility of a culture clash. The cultures of anestablished,stablecompanyandarisk-takingblitzscalerareaboutasdifferentascanbe.InitsbattleagainstAmazon,Walmartspent$3.3billiontoacquireJet.com,a
high price for a thirteen-month-old start-up (the price represented a revenuemultiple double that ofAmazon’s already-rich valuation). The two companieshavealreadyexperienced someculture clashmoments, suchaswhenWalmartaskedJettostopholdingitsregularofficehappyhours,stockingbottlesofliquorintheofficekitchen,andallowingemployeestodrinkattheirdesks.Accordingtoa2017WallStreet Journal article, Jet executivescomplained, andWalmartallowedJettoreviveitsofficehappyhours.On the other hand, Walmart’s e-commerce sales got a big boost from the
acquisition,andJethasallowedWalmarttoappealtourbanmillennials—akeydemographic that has typically shunned Walmart’s traditional stores.Blitzscaling is risky, but so is doing nothing if a competitor seems likely tosucceedinscalingup.
OPTION#3:AVOIDTHEM
The final and perhaps most often “successful” option is to cede the currentmarket to blitzscalers and use your current assets to migrate to a new, lessvulnerablemarket.Recallourlistof$100billiontechnologycompaniesfromtheintroduction; the oldest member of that list successfully pursued this exactstrategy.IBM was one of the original computer blitzscalers. IBM’s willingness to
invest in the growth of breakthrough products like the System/360mainframe
allowedittodominatecomputingfordecades.UndertheleadershipofThomasWatson Jr., IBM invested $5 billion to develop and launch System/360($30billionintoday’sdollars).ButbyApril1993,whenLouGerstnertookoverasCEO,IBMpostedan$8billionloss—thelargest inthehistoryofAmericanbusiness to that point—and seemed in danger of being surpassed by youngerblitzscalerslikeDell.Rather than ignore theproblemor try to competedirectly in thePCmarket
thatithadcreatedin1981,GerstnersuccessfullyrepositionedIBMasatrustedsystemintegratorandtechnologyconsultantforcorporateAmerica.ThescopeofIBM’smigrationcanbeseenintwotransactions:in2002,Gerstner’slastyearasCEO,IBMacquiredtheconsultingbusinessofPricewaterhouseCoopers,andin2005, it sold its personal computer business (including its iconic ThinkPadbrand) to a new blitzscaler from China named Lenovo (which also acquiredIBM’sserverbusinessin2014).Anotherpowerfulexamplecanbefoundinhowindependentbookstoreswere
abletoweathertheonslaughtfromAmazonandactuallymountacomeback.Noindependent bookstore can possibly compete with Amazon on availableselectionorprice.But thenumberof independentbookstoreshas increasedforeach of the last seven years, even asAmazon has continued to scale, becausethey’vemigratedoutofthebooksellingbusinessandintotheliterarycommunitybusiness, becoming destinations for cultural events like author signings, bookclubmeetings, spoken word performances, andmore. Independent bookstoresoffersomethingthatAmazoncannot(atleastuntilVRbecomesmoreadvanced):the experience of being in a bookstore, complete with the smell of books,friendlystaff,andthepresenceoffellowbibliophiles.Findingyourselfinthecrosshairsofablitzscalingcompetitorcanandshould
befrightening,butitisnotadeathsentenceifyouchoosetherightresponse.Butdecidequickly;thespeedofblitzscalingmeansthattakingyourtimeisthesameasdoingnothing.
PARTVI
ResponsibleBlitzscaling
In an idealworld,blitzscalingorganizationswouldembodyall thevirtues thatsocietymightdesire from itsbusinesses—adiverseand inclusiveworkforce, astrongsenseofresponsibilitytoshareholdersandstakeholders,anamplesupplyofwell-payingjobs,andexecutiveswhoserveasmoralrolemodelsandleadersof society. The unfortunate truth is that for all the good that blitzscalingproduces,blitzscalingorganizationscanbeguiltyofthesamesinscommittedbyothertypesofcompanies,andfacesomeinherentchallengesevenwhentryingtobehaveresponsibly.Blitzscalingcompaniesalmostalwaysoperateinfiercelycompetitivemarkets
where, inorder to surviveand thrive, theyneed tooutgrow their rivals. In thebest case, they do this by focusing relentlessly on building the businesswhilealsotryingtoachievebroadersocialgoals.Intheworstcase,theytrytogetbigfastbyanymeansnecessary.Thesepressuresarecompoundedbythefactthatblitzscalingcompaniesgrow
soquicklythattheyoftenbecomekeyplayersinsocietybeforethey’vehadtimeto fully mature. This can result in problematic corporate cultures, adversarialrelationshipswithregulators,andquestionabledecisionmaking.Thesechallengesare realbut shouldn’tdiscourageus fromblitzscaling.The
artliesinmarryingresponsibilityandvelocitysothatweareabletosuccessfullycapturethefirst-scaleradvantagewhilestilldevelopingandadheringtoastrongmoralcompass.Skeptics might argue that the kind of scale that blitzscaling produces is
inherentlybad,andthatsocietyshouldsimplypreventcompaniesfromgrowingtoobig.TestifyingbeforeCongressin1911,futureSupremeCourtjusticeLouis
Brandeis argued, “I thinkwe are in a position, after the experienceof the lasttwentyyears,tostatetwothings:Inthefirstplace,thatacorporationmaywellbetoolargetobethemostefficientinstrumentofproductionandofdistribution,and,inthesecondplace,whetherithasexceededthepointofgreatesteconomicefficiency or not, it may be too large to be tolerated among the people whodesiretobefree.”Wedisagreewith thispositionontheharmfulnessofscale in today’sworld.
First,Brandeiswas speakingduring theeraof“trusts,”when figures like J.P.Morgan consolidated American industry into powerful, giant companies likeU.S. Steel. But we believe that today’s blitzscalers are qualitatively differentthanGildedAgetrusts.Thosetrustsheldvirtualmonopoliesoverthesupplyofkey physical resources like steel and oil. Consumers had no alternatives andwere forced to do business with them. In contrast, companies like Apple andAmazonhavetowintheircustomerseveryday,andiftheyfail todoso,thoseconsumerscansimplybuyDelllaptopsandorderbooksfromBarnes&Noble.Second, we believe that while big can sometimes be bad, big can also be
great.Scalecreatesdominantcompanies,butscalealsocreatesenormousvalue.The smartphoneswe love, for example, aremass-market consumer electronicsthatdependoneconomiesofscale.WhileBrandeisisrightthatsocietyneedstopreventmonopoliesthatblocktechnologyorbusinessinnovationinthewaythatthe old AT&T monopoly suppressed the progress of telecommunications,today’s largestcompanieshaveactuallyenabled innovationand thecreationofeven more value by providing a platform for everything from businessproductivitysoftware(Slack)toentertainment(Netflix).Eventheconcentrationof capital that scale has produced isn’t all bad; it has allowed blitzscalers totackle “moonshots” like space travel (SpaceX) and autonomous vehicles(Google’sWaymo)thatmaydramaticallyimproveourlives.Asopposedtoreflexivelycallingforthebreakupofbigcompanies,thebetter
approachtotemperingthepotentialabusesofscaleistoleveragetheprinciplesfor a healthy republic that James Madison laid out in “Federalist No. 10.”Madisonwasaddressing thedangersof“factions;” that is, specificgroups thatactagainst the interestsof theentirecommunity.Madisonargued that factionswere a natural consequenceof liberty and, to safeguard against them, the beststrategywastocreateadiversesocietyinwhichnoparticularfactionwouldbeabletodominate.Madisonwrote,“Extendthesphere,andyoutakeinagreatervarietyofpartiesandinterests;youmakeitlessprobablethatamajorityofthewholewill have a commonmotive to invade the rights of other citizens; or if
such a commonmotive exists, it will be more difficult for all who feel it todiscover theirownstrength,and toact inunisonwitheachother.”Webelievethesameapproachappliestoeconomicsaswellaspolitics;inotherwords,thatagreatervarietyofpowerfulcompanies—iftheyarepreventedfromcolluding—cancounterbalancethemalevolentorselfishgoalsofanyoneparticularentity.It’s true that, as with anything in life, blitzscaling produces winners and
losers.Start-upscanandwillfail,andallentrepreneurialenterprisescreateriskfor founders, employees, and investors.At the same time, they also create thepossibility for new businesses, new innovations, and new jobs. But the mostsuccessful modern societies err on the side of freedom rather than trying tooutlaw all risks, and on the whole we are all better off because we allowentrepreneurstotakethoserisks.Italsotemptingtobelievethattheeasiestwaytoensureresponsiblebehavior
istolegislateit.Theproblemis,weliveinagloballycompetitivemarketplace.A government that slows the growth of companies within its borders byweighing them down with inflexible legislation is just making it easier forirresponsible blitzscalers from outside those borders to dominate emergingindustries.TaketheuproarthatensuedwhenitwasrevealedthatFacebookandTwitter
were exploited by parties—both foreign and domestic—to hack theAmericanelectionprocess.That’sclearlybad,andmeasuresshouldbetakentounderstandand address the vulnerabilities that left user data exposed. But imagine ifallthoseusershadinsteadadoptedasocialmediaplatformunderthejurisdictionof some other government? Most likely the American public wouldn’t haveknownabouttheissue,letalonehavetheabilitytoremedyit.The fact that Facebook is a global networkmakes it far easier for users to
connectwithpeoplefromaroundtheworld.Thereis,forexample,no“FacebookoftheUK.”ButFacebookisunderthejurisdictionoftheUnitedStates,nottheUK,whichmeansthatwhenBritishusers’datawascompromisedandaMemberofParliamentsentalettertoMarkZuckerbergrequestingthatheappearbeforeaParliamentarycommittee,Zuckerberghadnoobligation todoso.Suchare thelimitationsofregulatingbusinessesinaglobalizedworld.
BLITZSCALINGINSOCIETY
Responsible blitzscaling matters because successful blitzscalers often reach a
pointwheretheyaremorethanjustabusiness;theyactuallyaffectthefabricofthesocietyinwhichtheyoperate.SocialmedialikeFacebookandTwitterhavechangedhowweconsumeinformationandhowwecommunicate.MarketplaceslikeAlibaba and eBay provide economic opportunity—some dedicated sellerseven relyon them for their livelihoods.Sharingeconomyservices likeAirbnbcanbringmoretourismanddiversityintothecitiesinwhichtheyoperate.AndAmazon is changing the entire retail industry, which affects everyone. AsSpider-Manteachesus,withgreatpowercomesgreatresponsibility.We believe that the responsibilities of a blitzscaler go beyond simply
maximizing shareholdervaluewhileobeying the law;youarealso responsibleforhowtheactionsofyourbusinessimpactthelargersociety.Butevenbeyondthemoralimperatives,responsibleblitzscalingisgoodbusinessstrategy.Societyprovidestheecosysteminwhichyoulive,andinwhichyourbusinessoperates,whichmeans that it can rightly claim some responsibility for your success. Inotherwords,yoursuccessiscontingentuponsocietyfunctioningproperly.Herein Silicon Valley, some might fantasize about floating cities in internationalwaters,butthefactisthatblitzscalingbusinessesrelyontheruleoflaw,robustfinancialmarkets,andaneducationsystemthatproducestalentedemployeesanda healthy market of consumers. To paraphrase Warren Buffett, we win the“ovarianlottery”whenwe’rebornintoblitzscalingecosystems.Moreover,responsibleblitzscalingcanactuallyprotectagainstlegislationthat
threatens to slow growth trajectories. Regulation typically arises whengovernment believes that an industry isn’t behaving responsibly. For example,America(alongwithmanyothernations)hasenvironmentalregulationsbecausecompanieswereoncepollutingwithabandonandcausingharmtocitizensandnature. Smart blitzscalers realize that self-regulating can actually delay orpreempt government regulation. Entrepreneurs often complain that regulatorswritebadpolicybecausetheydon’tunderstandtheintricaciesofbusiness;self-regulationletsbusinessesapplytheirexpertisetofindingthemostcost-effectivewaystoachievesocialgoals.
FRAMEWORKFORRESPONSIBLEBLITZSCALING
The key to blitzscaling responsibly without sacrificing pace of growth isdeveloping the ability to distinguish between various forms of risk. Oursuggested framework for risk evaluation is to consider two separate axes:
KnownversusUnknownandSystemicversusNonsystemic.
Known Unknown
Systemic Known/Systemic Unknown/Systemic
Nonsystemic Known/Nonsystemic Unknown/Nonsystemic
Uncertaintybyitself isn’trisk; itsimplyproducesunknowns,andunknownsaren’t inherently negative. As anyone who has ever read a mystery novel ortraveledtoanewcityorlearnedanewlanguagecanattest,oneofthegreatjoysoflifeisthejourneyofdiscovery,ofturningtheunknownintotheknown.However, when you combine uncertainty with the possibility of a negative
outcome, you produce risk. The magnitude of the risk is a function of theprobability,andseverity,ofthatpotentialnegativeoutcome.Blitzscalingalwaysinvolves risks, but all risks aren’t equal. This is why you need to distinguishbetweensystemicandnonsystemicrisk.Nonsystemic risk is localized and, at most, affects a part of the system.
Systemicriskcanimpactorevendestroytheentiresystem,eitherdirectlyorastheresultofcascadingproblems.Forexample,thepossibilityofnuclearwarisaclear example of systemic—even extinction-level—risk. Even if we don’tbelievethatwecaneliminatethisriskentirely,themagnitudeoftheriskmakesitworthexpendingagreatdealofefforttoreducetheprobabilitythatitoccurs.Applying this analysis shows that a number of common fears about
blitzscaling are actually nonsystemic risks. For example, one common fear isthat blitzscaling will produce an oligarchy of powerful technology executiveswithtoomuchpoweroverourgovernmentandoursociety.Buteventoday,withtechnologyfirmsdominatingtheranksoftheworld’smostvaluablecompanies,traditionalbusinessmogulssuchasRupertMurdochandtheKochbrothershavehad a far greater influence over public policy than tech leaders such as JeffBezos,LarryPage,orMarkZuckerberg.An additional fear that is starting to be broadly voiced is that socialmedia
(largelyaproductofblitzscalingcompanies)isauniquelydangeroustechnologythat is harming consumers—especially the young—by addicting them andconsumingall theirattention.It iscertainlytruethatsomepeoplearespendingmoretimeproducingandconsumingsocialmediathanisoptimalfortheirhealthandproductivity.But is this reallyasystemicrisk?In2010,anarticle inSlateentitled “Don’t Touch That Dial!” enumerated themany times in history that
critics have argued that newmediums for consuming information would ruinsociety. Socrates warned against the pernicious effects of the written word,which he believed would harm memory. In the sixteenth century, ConradGessnertriedtocompilealistofeverybook,aneffortthatledhimtoconcludethatthenewfangledprintingpresshadresultedinanoverabundanceofdatathatwas “confusing and harmful” to the mind. The French statesman Guillaume-ChrétiendeLamoignondeMalesherbeswrotethatnewspaperssociallyisolatedtheir readers, who would otherwise get their news from their church pulpits.Despite these warnings, the written word, the printing press, and newspapershave brought tremendous benefits to humanity. It’s possible but unlikely thatsocialmediawillhaveaqualitativelydifferentimpactthananypreviousformofmedia, but we generally find that when people start saying, “This time it’sdifferent,”itusuallyisn’t.New technologies have always had the potential to lead to new problems.
Newspapersledtodemagogic“yellowjournalism.”Advertisingledtosnakeoilsalesmen. The answer wasn’t to ban newspapers or advertising, but to buildpoliciesandinstitutionstomitigatetherisksinvolved.That’swhywehavelibellawsandregulatorsliketheFCC.Andwithtime,audiencesthemselvesbecomemoresophisticatedanddeveloptheirown“immuneresponses.”Critics of socialmedia are correctwhen they point out the corrosive effect
socialmediahavehadonboththecivilityofpoliticaldiscourseandtheidealofobjective, evidence-based truth. These are real problems, and that means weshould try to fix them.Socialmedia should bemore transparent aboutwho ispaying for advertisements, and should require the same standards for truth inadvertisingasanyothermedium.On the other hand, there are technologies emerging from blitzscaling
companies that could pose real, systemic problems (yet get far less mediaattention).Syntheticbiology,drivenbyCRISPR-Cas9targetedgenomeediting,has the potential to produce huge benefits in medicine and agribusiness, butbrings with it the systemic risk of bad actors engineering a deadly globalpandemic.Changesanddevelopmentsinthisfieldhaveoccurredsoquicklythatitisdifficultforgovernmentstocreateintelligentregulatoryregimestomanagethese risks. Responsible blitzscalers should give serious considerations tosystemic risks and seek structural dialogue that involves a broad set ofstakeholders rather than defying or stonewalling regulators. Conversely,regulators shouldn’t assume that they know better than industry and makeunilateral decisions. Broad collaboration with transparency and open
communicationsisthebestwaytobothidentifythesystemicrisksandfigureoutthe least costly interventions to reduce them while still encouraging rapidinnovation.The systemic/nonsystemic distinction is dynamic, not static, andblitzscalers
should be prepared to change their approach accordingly. For example,Facebookhasbeenextensivelycriticizedforitsroleinthe2016USpresidentialelection, both for distributing deceptive content (aka “fake news”) and for notdoingenoughtoprotectitsusers’personaldatafrombeingexploitedbypoliticalconsulting firms likeCambridgeAnalytica.Bothof these issues are legitimateconcerns,sincetheybotherodethetrustthatFacebookusershaveinthecontenttheyfindonFacebookandinFacebookitself.Facebook’sscalehasmadeit thekeeperofvast trovesofdataonmorethan
200millionAmericans,aswellastheprimarywayinwhichmostAmericansgettheir news and share it with their friends. This means that the issues of dataprivacy and deceptive content not only affect Facebook and its users, but thefabricofsocietyitself.IfFacebookwerestillanichesocialnetworkforstudentsatIvyLeaguecolleges,theimpactwouldlargelybelocalized,butiftheseissuesdidinfactswaytheoutcomeofthe2016presidentialelection, thenthatwouldundoubtedlyrepresentasystemicjolt.In instances like these, a company may have to work together with the
governmentinordertoaddressaseriousissue.Incasesofsuchmagnitude,thedefaultimpulseisoftentocallforthecreationofanewregulatoryagency,butgovernment regulation alone has proven too slow to keep up with the rapidchanges of blitzscaling. At the same time, pure self-regulation hasn’t provensufficient. What’s needed is a dynamic public/private partnership wheregovernmentinputmesheswithprivateimplementation.Similarly, in thewake of all the revelations thatmisinformation spread via
social networks may have compromised the outcome of the election, theresponse of traditional media outlets like the New York Times and theWashingtonPostwastocallforFacebooktohirehumaneditorstopolice“fakenews.” This seems like a classic example of “When you have a hammer,everything looks like a nail.” You can’t simply apply the traditional editorialprocesses designed for a fifty-person newsroom to a platform with a billionpotential “reporters”writing billions of “articles” per day. Instead of trying tocopyandpasteasolution,Facebookshouldcomeupwithitsownideasforhowtoaddress theproblemand then find scalableways to implement them.These
solutions don’t need to be perfect; they just need to be better thanwhat camebefore, and importantly, continue to improveover time. Itwillbeachallenge,butwewouldn’tbesurprisedifthesolutionsultimatelyproduceafinalproductthat is even better than that of the old system, incorporating more voices,transparentfact-checking,andsocialproof.
THERESPONSESPECTRUM
Onceyouhavecategorizedariskasknownversusunknownorsystemicversusnonsystemic,youneedtodecidehowyouwillrespond.Webelievethatpotentialresponsesfallintofourbroadcategories.
#1:TAKEDECISIVEACTIONNOW.
Systemicrisksmayrequireanimmediate,“stopthepresses”response.In2011,forexample,anAirbnbhostinSanFranciscocamehomeanddiscoveredthatanAirbnb guest had trashed her house and stolen her possessions, including hergrandmother’sjewelry.Airbnb’sinitialresponse,whichwastocoordinatewiththepolicedepartmentandcompensatethehostfinanciallybuttoemphasizethatsuch incidents would be dealt with on a case-by-case basis, may have beenlegally sound, but didn’t address the systemic issue—hosts losing trust inAirbnb.After he recognized the magnitude of the problem, Brian Chesky took
decisiveaction.First,heaccepted full responsibility, inwriting,on theofficialAirbnb blog, “With regards to EJ, we let her down, and for thatwe are verysorry.We should have responded faster, communicatedmore sensitively, andtakenmoredecisiveactiontomakesureshefeltsafeandsecure.Butweweren’tprepared for the crisis and we dropped the ball.” Second, he announced theAirbnb Guarantee, whereby the company would protect hosts against up to$50,000inpropertydamage.Theseactionswereabsolutelynecessarygiventhescope and potential impact of the crisis, not just forAirbnb but for thewholeindustry. (You can readBrian’s full response, “OurCommitment toTrust andSafety,”ontheofficialAirbnbblog.)
#2:TAKESHORT-TERMACTIONNOW,BUTDEFERPERMANENT
ACTIONUNTILLATER.
Evenifariskissystemic,itmaybepossibletoemployashort-termpatchthatcan later be replaced by a permanent fix. At PayPal, credit card fraud wasdefinitely a systemic and existential issue. After all, a payments system thatusers don’t trust is worthless. But we didn’t have an immediate solution forpreventingsuchfraud.Soourresponsewastoeatthecostsourselvessothatourusersweren’taffected.Weknewthiswasatemporarysolution,butitboughtusthetimeweneededtobuildstrongerfrauddetectionintotheproduct.
#3:NOTETHEPROBLEMNOW,ANDCOMMITTOTAKINGACTIONLATER.
Iftheriskismanageablenowbutwillbecomesystemicinthefuture,youcan’tsimply ignore the problem.Even if you don’t take any immediate action, youshouldcommittoactionlatersothatwhentheriskdoesbecomesystemic,youaren’tcaughtoffguard.IntheearlydaysofPayPal,inadditiontotheproblemofcreditcardfraud,we
also faced the issue of illegal transactions. We obviously didn’t want peopleusing PayPal for buying and selling drugs or funding criminals and terrorists,whichwould represent a systemic risk.On the other hand,we didn’t have in-houseexpertise in forensicaccountingorpolicework.Becauseour transactionvolume was still low, and because we judged the probability of illegaltransactionsoccurringtobeverylow,wedeferredworkingontheproblem,butwe also committed to building the necessary expertise and infrastructure tobettermanagetheissuelateron.
#4:LETITBURN.
Whenyou are facing anunknown/nonsystemic risk, itmaynot evenbeworthexpendingtheefforttoanalyzeit—it’sprobablyasmallfirethatyoushouldletburn.
BALANCINGRESPONSIBILITYANDVELOCITYASTHEORGANIZATIONGROWS
Balancingthedualprioritiesofresponsibilityandvelocityisatrickydancethatmay look very different at each stage of growth.We’ve observed some broadpatternsthatseemtoapplytomostcompanies.Earlyon,duringtheFamilyandTribestages,responsibleblitzscalingmeans
clearlydefining thecompany’smissionand laying thefoundationforaculturethat values being a responsible part of a larger society. To do so, you shouldimagine a future in which the company has succeeded in becoming a globalgiant, and then evaluate the likely impact of that success on your keystakeholdersandonsocietyasawhole.For example, does your company produce negative externalities in which
transactionsbetweenyouandyourcustomersimposecostsonexternalparties?John D. Rockefeller might not have realized the impact that blitzscalingStandardOilwouldultimatelyhaveon theglobal climate,buthisdescendantsseem to have done so, given that in 2016 the Rockefeller Family Fundannounced that it would immediately divest its holdings in ExxonMobil…thelargestcorporatedescendantofStandardOil.Ideally,youwanttopredicttheseexternalities while you still have time to either radically reshape the businessmodel or simply get into another business, since it’s easier to institute radicalchangeorabandontheprojectaltogetherwhenyou’restillverysmall.Atthisstageyoushouldalsotakeactionsthatanticipatetheinternaleffectsof
growth. For example, blitzscaling companies need to hire so quickly that theyoftenrelyonpersonalnetworkstosourcejobcandidates.Appliedcarelessly,thistechniquecanresultinahomogenousandnoninclusiveculture.Butifyoubuilda diverse and inclusive network before you scale, hiring within the networkdoesn’tposeasmanydiversitychallengeslateron.AsthecompanyachievessuccessandgrowsintotheVillagestage,it’stimeto
ask yourself, “What things, if I don’t fix them now, will be functionallyimpossible tofixatscale?”It’sespeciallydifficult tofind thebalancebetweenmoralityandvelocityduringthisstage,becausethecompanyisprobablyfiringonallcylindersandpursuingall-out lightning-fastgrowth,and ifyoupauseorslowdowntofixthings,acompetitormightgrabthefirst-scaleradvantagefromrightunderyournose.That’swhy thequestion askswhat is “impossible,”notjustwhatis“difficult.”You should also continue to give serious thought to the potential negative
impactofyoursuccess.Inearlierstages,youweresimplyspeculatingaboutthefuture;bytheVillagestage,youhaveenoughdatatoextrapolateintothefuture
with reasonable accuracy.Youmight still bewrong, but if you don’t performthis exercise you’ll be culpably negligent when it comes to your moralobligationsiftheworsthappens.OnceyourcompanyreachestheCityorNationstage,itnowneedstotakeon
the responsibilities of an incumbent, which are very different from theresponsibilities of a challenger. Remember when you asked yourself whichproblemsyoucouldfixlater?Well,laterjustarrived.Ifyoupreviouslyignoredissues such as diversity, legal compliance, or social justice, you need tounderstandthatalleyesarenowonyou,andyou’llbeexpectedtobehaveasaresponsiblecitizenandrolemodel.Plus,ifyoudon’ttackletheseresponsibilitiesproactively,you’llhave to tackle themreactively—whichwillalmostcertainlybemorecostlyandmorepainful.Likeitornot,whenyourcompanyisaCityoraNation,youneedtostartthinkinglikeamayororapresidentandsetrulesforthegoodofhumanityasawholeratherthanjustforthegoodofyourprofits.
CONCLUSION
Over the past few decades, blitzscaling has redefined countless industries andhelpedshapenearlyeverypartofourlives.Eachwakinghour,youprobablyusemultiple products from companies that blitzscaled in the past or are in theprocessofblitzscalingtoday.ButwhatiftheBlitzscalingEraisjustgettingstarted?Sofar,blitzscalinghas
been concentrated in software and the Internet, but it’s likely to reshape ourphysical infrastructure or even our bodies in the future. Artificial intelligencewill soon be ubiquitous, thanks to self-driving vehicles and better machinelearning. Technology innovations in the life sciences, such as CRISPR geneediting, may change the fabric of life itself. Cryptocurrencies and blockchaintechnology may change the role of governments and corporations in globalfinanceandcommerce.Newtechnologiesareemerging rapidlyandpromise tochangeeverything—
again.Thesenewtechnologieswillenablenewbusinessmodels,which in turnwillcreatenewindustries.Inthehistoryofhightech,platformshifts,suchasthemove frommainframes to client-server or themove from theWeb tomobile,haverepresentedhugeopportunities.Today,multipleplatformsareemergingorshiftingsimultaneously,bringinggreatercomplexity—andevengreaterrewardsforspeed.Meanwhile,marketsandinvestorsareincreasinglywillingtofundaggressive
bets on blitzscaling. Because private investors are willing to fund growth,companiesarestayingprivatelongersothattheycankeepmakinginvestmentsinblitzscalingthatthepublicmarketsmightfrownupon.CompanieslikeAirbnbandXiaomihavevaluationsinthetensofbillions,makingthemmorevaluable(on paper) than the vast majority of publicly traded companies. Becauseinvestors in the public markets aren’t able to profit as much from post-IPOblitzscaling, those investors are looking to invest in privately held companies,whichmakesevenmoremoneyavailabletofundblitzscaling!
In this book, we’ve tried to help the various stakeholders in society betterunderstand the phenomenon of blitzscaling, how it’s changing the world, andhowtorespondtoit.Entrepreneursshouldbeawarethatblitzscalingisthemainpatternbywhich
major new technologies, business ecosystems, and companies establishthemselvesand replace theirpredecessors.With theknowledge theygain fromthisbook,entrepreneurscanbetterapplyitsmethodstotheirownbusinesses,bemore aware of how their competitors might employ the same techniques tochangetheplayingfield,andbebetterpreparedtorespondtothosecompetitivethreats.Theywillalsobetterunderstandhowtoblitzscaleresponsibly,andbuildcompaniesthatimprovesocietyandofwhichtheycanbeproud.Corporate executives and organizational leaders need to recognize that
blitzscalingislikelytoimpacttheirindustriesandbusinessessoonerratherthanlater.Becausetechnologyisbecomingsointegraltoeverybusiness—remember,allcompaniesarebecomingtechcompanies—thespeedoftechnologicalchangeisincreasingthespeedofchangeforeverybusiness.Understanding blitzscaling allows established businesses to better anticipate
andadapt tochangesinthemarket landscape.Somechangesmightblowover.But others will change everything, and require everyone, including marketleaders, to change accordingly. Adapting is seldom easy for big companies;everything from capital structure to organizational incentivesmake it difficultfor them to take big risks.But themarket leaderswho use the lessons of thisbook to defend themselves against blitzscaling competitors while investing inblitzscaling their own new businesses will be the ones who remain marketleadersinthefuture.Governments, politicians, and regulators should try to understand how
blitzscalingcanhelpratherthanharmsociety.Therapidchangethatblitzscalingbrings can be disruptive and thus frightening. The natural impulse is to try toslowdownblitzscaling,whetherthroughtaxesorregulations.Theproblemwithgiving in to this understandable instinct is that change is going to happenwhetheritoriginatesinyourbackyardornot.Slowingthingsdownmightmakeyoufeelmorecomfortable,butitcomesatthecostofallowingcompetitorsfromotherareastogainlastingdominanceoftheglobalmarket.Blitzscalingattractsinvestmentcapitalandcreatesmajornewindustries;asacommunityoranation,youwantmoreblitzscalingcompanies,notfewer.Abetterunderstandingofthepositivesandnegativesofblitzscalingwillhelp
governmentsnotonly tomake theappropriateadjustments toencourage it,butalsotoimprovethechancesofarrivingattherightsocialoutcomes.TheeconomicreformandgrowthofChinaoverthepastthirtyyearshaslifted
eight hundred million people out of poverty—more than any other policy orprogramduringthattime.Despitetheveryrealsocialandenvironmentalpriceofthatgrowth,theworldismuchbetteroffforit.Blitzscalingalsoimprovessocialmobility.ComparedwithachildborntoparentsinDetroit’spoorest20percent,a child born to parents in San Francisco’s poorest 20 percent has double thechance of ending up in the richest 20 percent as an adult. We believe thatblitzscalingcanbringthatkindofeconomicmiracletootherareasoftheworld,and that educated blitzscalers will be more likely to fulfill their ethicalobligationstostriveforpositivesocietalimpact.Consider,forexample,thepositiveimpactthatthemobilebankingserviceM-
Pesa has had in Africa since its introduction in 2007. It has raised incomes,boostedeconomicgrowth,andfinanciallyempoweredwomen.WhenAlexanderHamilton proposed a nationwide banking system for the United States in the1790s, it took nearly a century for his vision to be realized. Thanks toblitzscaling,M-Pesadidthisformultiplecountriesinjusttenyears.Progress occurswhennew ideas emerge and spread.Sometimes these ideas
taketheformoftechnologiesliketheprintingpressorthesmartphone,andothertimestheyremainabstract,likedemocracyorcapitalism.Blitzscalingmaybeanabstractmeme,butithashadaveryconcreteimpactontheworld.Thememeofblitzscalinggot its start inSiliconValley, took root inChina,and is spreadingquickly—theonlywaythatblitzscalingknowshow.Asitspreads,italsoactsasacatalyst,helpingtoacceleratetheimpactofotherideas.Wewouldlovetoseethisbookhelptotransformeveryregion—Africa,theMiddleEast,Europe,LatinAmerica,aswellasNorthAmericaandAsia(wheretheUnitedStatesandChinahaveledtheway).Here’swhatallofusneedtorealizeabouttheBlitzscalingEra:Speedanduncertaintyarethenewstability.Theonlywaytothriveinthisfast-changingworldistoaccepttheinevitability
ofchange.Useittoyouradvantage,whetheryou’refocusedonyourindividuallifeorthefateofanation.This book is actually the third in a series that covers adapting to the
NetworkedAge.TheStartupofYoufocusesonhowindividualscanadapttheircareerstoarapidlychangingworldbyremaininginastateof“PermanentBeta.”
(Visit thestartupofyou.com for more resources and inspiration.) The Allianceanalyzes how companies and managers should adapt their talent-managementstrategies to build stronger relationships with employees despite an uncertainfuture.(Visitalliedtalent.comtogethelpintroducingtheseframeworksintoyourorganization.) This volume is both a prequel and a sequel; it explains howblitzscalinghelpedcreate theNetworkedAge, andhowentrepreneurs, leaders,companies,andgovernmentscanshapethecomingchange.First,beaninfinitelearner.Thebestandworstthingabouttherapidpaceof
changetodayisthattherearenoexpertswithten-plusyearsofexperienceinanyemerging phenomenon. If you’re able to climb the learning curve faster thanothers,youhavetheopportunitytocreatemassivevaluefromit.Whilewewishwecouldwriteasimple,comprehensivelistofrulesthatwouldguaranteeyoursuccess,it’sunclearhowanyonecoulddescribeastrategythatwouldapplytoallthepotentialchangesthatwilloccurinthenextfewyears,letalonedecades.Thelandscapeisalwayschanging,andlearningishowyouadapt.Second, be a first responder. As new technologies and trends emerge, the
uncertaintyofwheretheyareheadedwillparalyzemanypeopleandkeepthemfrom acting. Those who are willing to act—and act quickly—despite theuncertainty will have a disproportionate advantage. Seek out blitzscalingcompanies and markets; that’s where you’ll find the greatest growth andopportunity.Finally, and somewhat paradoxically, be a sourceof stability. In aworldof
constant change and uncertainty, people will need reassurance and support.Offeringstabilityandcalminthemiddleofthestormwhileothersarecaughtinthetumultwillmakeyouanaturalleader.This prescription may seem scary, but we believe that this era of intense
competition can be a good thing. Competition may be challenging for theindividualpersonorcompany,but it isgoodfor thecollectivewhole.Asmoreregions and ecosystems promote blitzscaling, more net value will be created.Like biodiversity, this “blitz-diversity” supports different types of growth andwill allowblitzscaling tobe applied to abroader arrayof importantproblems.Blitzscalingalsohelpsguardagainststasisandcomplacency,because itallowsnewdomainstoemergeandgrowquickly,forcingincumbentstoadapt.Ifyoubelievethefuturewillbebetterthanthepast,blitzscalingisheartening,
becausewe’llget therefaster. Ifyoubelieve that thefuturewillbeworse thanthepast,blitzscalingisterrifying,becauseitoverturnstheexistingorderfaster.
Hereishowwepersonallyfeelaboutblitzscaling:Webelieve that the future can and shouldbebetter than thepast, and
thatit’sworthtoleratingthediscomfortwefeelwhenblitzscalingtogettothefutureasquicklyaswecan.Wehopetoseeblitzscalingenablemoreentrepreneurstobuildtransformative
companiesandsucceedatamassivescale.We hope to see more established companies leveraging the lessons of
blitzscalingtobemoreadaptableandbetterequippedtotacklethechallengesofthefuture.We hope to see activists and governments use the tools of blitzscaling to
changetheworldforthebetter.Thecompaniesthatchoosetoblitzscalewillsoonsetthepaceofprogressin
every industry. It’s up to you to lead this change—for yourself, for yourcompany,andforsocietyasawhole.Raceyoutothefuture.
ACKNOWLEDGMENTS
Thankstoourfamiliesfortheirsupportandpatiencethroughthislongprocess—Michelle,Alisha,Jason,andMarissa.Thankyoutooureditor,TaliaKrohn,andhercolleaguesatCurrencyforgivingahometoourideas.LisaDiMona,MeganCasey,DavidSanford,SaidaSapieva,BrettBolkowy,andIanAlasonourteamofferedcriticalsupportthroughoutthisjourney.Mehran Sahami, sponsored ourCS183C class at Stanford,whichwe taught
withourfriendsandfellowinstructorsAllenBlueandJohnLilly.Thanksaswelltotheguestswhosharedtheirstorieswiththeclass,manyofwhichmadeitintothe book, including Sam Altman, Brian Chesky, Patrick Collison, MichaelDeering,DianeGreene,ReedHastings,MarissaMayer,ShishirMehrotra,AnnMiura-Ko,MariamNaficy,JenniferPahlka,EricSchmidt,SelinaTobaccowala,NiravTolia,andJeffWeiner.ManythankstoeveryoneatGreylockPartnersfortheirsupportofthisproject,
including Joseph Ansanelli, Jerry Chen, Josh Elman, Chris McCann, StaceyNgo,SimonRothman,andElisaSchreiber.WeowemuchtoJuneCohen,DeronTriff,andtherestoftheWaitWhatteam,
whoproducetheMastersofScalepodcast.ManyofthestoriesinthebookcomefromtheguestswhoappearedinSeason1andSeason2,includingAneelBhusri,SaraBlakely,StewartButterfield,BarryDiller,JohnElkann,CaterinaFake,TimFerriss, PayalKadakia,NancyLublin,Mark Pincus, LindaRottenberg, SherylSandberg,HowardSchultz,PeterThiel,TristanWalker,EvWilliams,andMarkZuckerberg.Thanksalsotothosewhomadeacameoonthoseepisodes,includingUmber
Ahmad,DominiqueAnsel,GregBaldwin,AlexaChriston,PauletteMaeCole,Chris Costa, Lisa Curtis, Susan Danziger, Angela Duckworth, Kara Goldin,Natasha Hastings, Margaret Heffernan, Drew Houston, Joi Ito, Leila Janah,DanielKahneman,CherylKellond,DaraKhosrowshahi,JoshKopelman,OmidKordestani, Michelle Lee, Tim Lefler, Kristen Marhaver, Kathryn Minshew,
AndrewNg,AubriePagano,HadiPartovi,RobertPasin,JulianaRotich,AndrésRuzo,DickStockton,TonyTjan,YossiVardi,andDarrylWoodson.TheauthorsoftheSiliconGuildprovidedvaluablefeedbackonearlierdrafts,
including Peter Sims, Jennifer Aaker, Nancy Duarte, Morten Hansen, FransJohansson,CharleneLi,TinaSeelig,ChrisShipley,Anne-MarieSlaughter,andCarolineWebb.Somanyothershelpedusalongtheway,includingBenCasnocha,EladGil,
Bing Gordon, Fred Kofman, Dmitri Mehlhorn, Marten Mickos, ChristopherSchroeder,MikeVolpi,andPatWadors.AndthankyoutoBillGatesforgenerouslytakingthetimetocontributethe
forewordtothisbook.
APPENDIXA:DISCLOSURES
As an entrepreneur and investor, either Reid Hoffman or the venture firmGreylock Partners, where he is a general partner, have the followingrelationshipswithcompaniesmentionedinthisbook:
Airbnb:Greylockportfoliocompany;investorandboardobserverCloudera:GreylockportfoliocompanyDropbox:GreylockportfoliocompanyFacebook:Greylockportfoliocompany;personalinvestmentFriendster:personalinvestmentGladly:GreylockportfoliocompanyGreylockPartners:generalpartnerGroupon:GreylockportfoliocompanyInstagram:GreylockportfoliocompanyLinkedIn:cofounder,GreylockportfoliocompanyMedium:GreylockportfoliocompanyMicrosoft:boardmemberMozilla:formerboardmemberNextdoor:GreylockportfoliocompanyPandora:GreylockportfoliocompanyPayPal:foundingboardmemberandexecutivePureStorage:GreylockportfoliocompanyRedHat:GreylockportfoliocompanySocialNet:cofounderTumblr:GreylockportfoliocompanyZynga:formerboardmember;personalinvestment
APPENDIXB:THEBLITZSCALERS
Throughout thisbook,we tell thestoriesofvariousblitzscalers.Thisappendixincludesbriefprofilesthatprovidebasiccontextforthecuriousreader.
AIRBNBAirbnb.com
Airbnbisanonlinemarketplaceandhospitalityservice,enablingpeopletoleaseorrentshort-termlodgingincluding vacation rentals, apartment rentals, homestays, hostel beds, or hotel rooms. FoundedAugust2008,SanFrancisco,CA
ALIBABAAlibaba.com
The Alibaba Group is an e-commerce, retail, and technology conglomerate that provides consumer-to-consumer, business-to-consumer, and business-to-business services including electronic payments andcloudcomputing.FoundedApril1999,Hangzhou,China
AMAZONAmazon.com
Amazonisane-commercecompanythatalsoproducesconsumerelectronicsliketheKindleandEchoandistheworld’slargestproviderofcloudcomputingservices.FoundedJuly1994,Seattle,WA
APPLEApple.com
Appledesigns,develops,andsellsconsumerelectronics,computersoftware,andonlineservices, suchastheiPhone,iOSoperatingsystem,andMacpersonalcomputers.FoundedApril1976,LosAltos,CA
CHARITY:WATER
Charitywater.org
Charity:Water is a not-for-profit organization that provides clean and safe drinking water to people indevelopingnations.FoundedSummer2006,NewYork,NY
CHESAPEAKEENERGYChk.com
Chesapeake Energy is a petroleum and natural gas exploration and production company. FoundedMay1989,OklahomaCity,OK
CLASSPASSClassPass.com
ClassPass offers a flat-rate monthly subscription service that allows subscribers to attend participatingfitnessclassesaroundtheworld.FoundedJune2013,NewYork,NY
DRESSFORSUCCESSDressforsuccess.org
DressforSuccessisanot-for-profitorganizationthatprovidesanetworkofsupport,professionalattire,andthedevelopmenttoolstohelpwomenthriveinworkandinlife.Founded1997,NewYork,NY
DROPBOXDropbox.com
Dropboxisafilehostingservicethatofferscloudstorage,filesynchronization,personalcloud,andclientsoftware.Founded2007,MountainView,CA
FACEBOOKFacebook.com
Facebook provides products such as Facebook, Instagram, andWhatsapp that enable people to connect,share,discover,andcommunicatewitheachother.FoundedFebruary2004,Cambridge,MA
FLIPKARTFlipkart.com
Flipkart is an e-commerce company that focuses on serving the India market. Founded October 2007,Bangalore,India
GOOGLEGoogle.com
AlphabetInc.isaholdingcompanythatincludesGoogle(thecompany’scoreinternetbusinesses),aswellasothernon-internetcompaniessuchasCalico,Verily,Waymo,X,andNestLabs.Inthisbook,werefertothecompanyasGoogle,bothbecauseitisthenamebywhichmostpeopleknowthecompany,andbecausewefocusonthecompany’sInternetbusinesses.FoundedSeptember1998,PaloAlto,CA
GROUPONGroupon.com
Grouponisane-commercemarketplacethatconnectsitssubscriberswithoffersfromlocalmerchants.Itsprimaryfocusareasareactivities,travel,goods,andservices.FoundedJanuary2008,Chicago,IL
KHANACADEMYKhanacademy.org
KhanAcademy’smissionistoprovideafree,world-classeducationforanyone,anywhere.Itdoesthisbyofferingonlinepracticeexercisesandinstructionalvideos.FoundedOctober2006,MountainView,CA
LINKEDINLinkedIn.com
LinkedInistheworld’slargestprofessionalnetworkandseekstoconnecttheworld’sprofessionalstomakethemmoreproductiveandsuccessful.FoundedDecember2002,MountainView,CA
MERCADOLIBREMercadoLibre.com
MercadoLibreprovidessolutionstoindividualsandcompaniesbuying,selling,advertising,andpayingforgoodsonline.FoundedMay1999,BuenosAires,Argentina/Stanford,CA
MICROSOFTMicrosoft.com
Microsoft develops,manufactures, licenses, supports, and sells computer software, consumer electronics,personal computers, and services. Based on revenue, it is theworld’s largest computer softwaremaker.FoundedApril1975,Albuquerque,NM
M-PESAvodafone.com/content/index/what/m-pesa.html
M-Pesa is amobile phone-basedmoney transfer, financing, andmicrofinancing service that launched inKenyabutservesmarketsaroundtheworld.FoundedMarch2007,Nairobi,Kenya
NETFLIXNetflix.com
Netflix is an Internet entertainment service that offers its members TV shows and movies, includingoriginal series, documentaries, and feature films. Members can watch as much as they want, anytime,anywhere,withnocommercialinterruptions.FoundedAugust1997,ScottsValley,CA
PAYPALPayPal.com
PayPaloperatesaworldwideonlinepaymentssystemthatsupportsonlinemoneytransfersandservesasanelectronicalternativetotraditionalpapermethodslikechecksandmoneyorders.FoundedDecember1998,PaloAlto,CA
PRICELINEPriceline.com
Pricelineprovidesonlinetravelandrelatedservicestoconsumersandlocalpartners.ItsprimarybrandsareBooking.com, priceline.com, agoda.com, KAYAK, Rentalcars.com, and OpenTable. Founded 1997,Stamford,CT
ROCKETMORTGAGERocketMortgage.com
ThroughtheRocketMortgagewebsiteormobileapp,userscanuploadfinancialdetailsandgetamortgageloandecisioninminutes.QuickenLoanslaunchedRocketMortgageinNovember2015,Detroit,MI
SALESFORCE.comSalesforce.com
Salesforce.com provides cloud-based applications for sales, service, and marketing, as well as enablingpartners to offer and run their own solutions on the Salesforce Platform. Founded February 1999, SanFrancisco,CA
SLACKSlack.com
Slackprovidescloud-basedcollaborationtoolsandservicesthatconnectteamswiththeapps,services,andresourcestheyneedtogetworkdone.Founded2009,Vancouver,BritishColumbia,Canada
SPOTIFYSpotify.com
Spotifyisamusicandpodcaststreamingservicethatallowsuserstocreateandlistentoplaylistsaswellasindividualtracks.FoundedApril2006,Stockholm,Sweden
STRIPEStripe.com
Stripehelpsbusinessesacceptpaymentsonlineandinmobileapps.Founded2010,PaloAlto,CA
TENCENTTencent.com
Tencent isaholdingcompanywhosesubsidiariesprovidevarious Internet-relatedservices,products,andtechnology both in China and globally. Its major services include QQ and WeChat. FoundedNovember1998,Shenzhen,China
TESLATesla.com
Tesla is an automaker, energy storage company, and solar panelmanufacturer. Founded July 2003, SanCarlos,CA
TWITTERTwitter.com
Twitterisanonlinenewsandsocialnetworkingservicewhereuserspostandinteractwithmessagescalled“tweets.”FoundedMarch2006,SanFrancisco,CA
UBERUber.com
Uberisatransportationtechnologycompany.Itdevelops,markets,andoperatestheUbercartransportationandfooddeliverymobileapps.FoundedMarch2009,SanFrancisco,CA
XIAOMIMi.com
Xiaomiisaelectronicsandsoftwarecompanythatdesigns,develops,andsellssmartphones,mobileapps,laptops,andrelatedconsumerelectronics.FoundedApril2010,Beijing,China
ZARAZara.com
Zara(anditsholdingcompany,Inditex)istheworld’slargestclothingandfashionretailer.FoundedMay1974,Arteixo,Spain
APPENDIXC:CS183CESSAYS
Oneofthewayswedevelopedthematerialinthisbookwasbyteachingaclassat Stanford University in the fall of 2015. This class, CS183C: Technology-enabled Blitzscaling, helped us refine our ideas and provided some of thecontentintheformofquotesfromthevariousdistinguishedguestswhovisitedourclass.Duringthecourseofourclass,weaskedourstudentstowritetwoessaysand
afinalreflection.We’veincludedlinkstothebestoftheessays,andasampleofthefinalreflections,bothtorewardourstudentsfortheirhardwork,andtogiveyou, the reader, some additional perspectives on the topic of blitzscaling. Forthoseofyouwhoare reading thepaperversionof thisbook,youcan find thelinksbelowontheBlitzscaling.comwebsite.
THEBESTOFESSAY1medium.com/cs183c-blitzscaling-class-collection/featured-essays-for-assignment-1-f8b34938e5e2
OguzhanAtayRobertChunJorgeCuetoAxelEricssonJocelynNeff
THEBESTOFESSAY2medium.com/cs183c-blitzscaling-class-collection/featured-essays-for-assignment-2-c620149f8eb5
JorgeCuetoSkylarDorosinAaronKalbJocelynNeff
ASAMPLEOFFINALREFLECTIONSChaitanyaAsawa:medium.com/@casawa/ride-of-your-life-
678bea009d3fChristinaChen:medium.com/@christina.chen/teachers-open-the-door-
you-enter-by-yourself-c9135aadef92JorgeCueto:medium.com/@jcueto/taking-the-leap-399ec46cf3a5MaxineCunningham:medium.com/@mmcunnin/blitzscaling-with-reid-
hoffman-co-final-assignment-62e921ba2bf3SkylarDorosin:medium.com/@sdorosin/from-household-tonation-final-
musings-on-blitzscaling-2b8b6e27a3ceAxelEricsson:medium.com/@ericsson_axel/lightning-fast-final-essay-
on-blitzscaling-612d12fc2139AndreEsteva:medium.com/@andreesteva/cs-183c-final-essay-
blitzscaling-a-foundation-for-rapid-company-growth-e59043d63292VijayGoel:medium.com/@vijaygoel/blitzscaling-knowing-when-it-s-
time-to-go-all-in-55f4cad85aaaMarcusGomez:medium.com/@mvgomez/final-lessons-6ac03fdb1397RishGupta:medium.com/@rish_says/what-i-learnt-from-reid-hoffman-
brian-chesky-marissa-mayer-elizabeth-holmes-jeff-weiner-on-1e66bf61a23a
KurtHeinrich:medium.com/@kurtjheinrich/cs183c-blitzscaling-takeaways-final-essay-10609b080562
BrandonHill:medium.com/@brandon_hill/how-and-when-to-blitzscale-f54c31f2a4fd
TeddyJungreis:medium.com/@teddyjungreis/blitzscaling-for-dummies-c3b48272acec
AaronKalb:medium.com/@kalb/blitzscaling-retrospective-b8e72bf81229
DanielKharitonov:medium.com/@volkfox/cs183c-final-essay-1a3242eca9f
CharlesLu:medium.com/@charleslu/like-lightning-638c9051beb8RyanMcKinney:medium.com/@ryanmckinney/blitzscaling-the-future-
8c9c27c1e1e7JoannMcMaster:medium.com/@joannmacmaster/99c620beaa8a
JocelynNeff:medium.com/cs183c-blitzscaling-student-collection/blitzscaling-a-chemical-reaction-bf9e318fe903
NirmitParikh:medium.com/@Nirmit_Parikh/cs-183c-blitzscaling-168d208532aa
VeeralPatel:medium.com/@vral/2ab47a57a162DayneRathbone:medium.com/@daynerathbone/blitzscaling-takeaways-
73570800f84bShikharShrestha:medium.com/@shikharshrestha/final-reflections-on-
blitzscaling-a8eb5aacba96JasonWeeks:medium.com/@Weeksy_J/cs183c-final-assignment-
9be1b4af8087
What’snextonyourreadinglist?
Discoveryournextgreatread!
Getpersonalizedbookpicksandup-to-datenewsaboutthisauthor.
Signupnow.