metrics drive design - eewei chen
DESCRIPTION
Presented at the UXPA September event on User-Centred Design and AgileTRANSCRIPT
[email protected] / @emotivu / Emoti.vu
Metrics Driven DesignBetter define success, deliver impact, make money
Eewei Chen@Ultraman
http://www.ml.sun.ac.za/wp-content/uploads/2009/08/social_graph.png
“Efforts and courage are not enough without purpose and direction.”John F. Kennedy
20 years
What is success?
Stories, workflows, wireframes, interactions
My Flights
ALERT:Manila to Cebu
Book A Flight
Get The Best Exchange Rates
Flight Status
CP100 - OK
Check-in
Available (1)
Destination Guides CEB Club
Commercial Partners About Cebu Pacific
Sign-inRegister
Shake Iphone to view more Special Flight Offers
MANILA TO
BORACAYOne way from
P780
DISRUPTION: Manila to Cebu delayed
Abstract Concrete
UX Assets?
Working software
Burn down
Collaboration
Refactoring
Agile
good enough
Disruption
no longer exist
no one can live without
Survival?
Can UX success be measured?
Pack selection
+ 213%Checkout completion
+ 252%User satisfaction*
+ 237%
Sky Upgrades
Free Trial uptake
+ 137%Subscriptions uptake
+ 187%User satisfaction*
+ 223%
Now TV Subscriptions
Query resolution
+ 120%Info retrieval speed
+ 239%User satisfaction*
+ 174%
Sky.com Nav
Problem: Poor trial and subs uptake
Solution: • New marketing animation• More effective brand messages• Timed marketing triggers• Remote usability test (2K users)• 4 week POC, 12 weeks
Problem: Too many choices, low upgrade conversion
Solution: • Usage based recommendations • 9 packs down to 3• Quick check out• MVT Testing to 1% traffic• 3 week POC, 8 weeks delivery
Problem: Inaccurate FAQs and MySKY tasks
Solution: • Super nav drop down• Quick tasks• Simplified IA• Remote usability test (2K users)• 2 week POC, 6 weeks delivery
*In App customer survey feedback form
BUT what if there isn’t much to go on?
(Case study)
Digital Airline Experiences are fragmented
Fear & Punishment works??
Apple wants it all
Many platforms
Stay connected with travellers
• Data: Passenger info• Need: More connected mobile digital experience• ROI: Mobile Ad wallet share• UX: Useful traveller engagement throughout lifecycle
Airline
Airline
Airline
Advertiser
Advertiser
Advertiser
Opportunity
Mobile Advertising
US$11.4 Billion (i) Mobile Ad spend in 2013
£500 million UK spend in 2012 is most per user (ii)
(i) -‐ h'p://www.gartner.com/newsroom/id/2306215(ii) -‐ h'p://www.guardian.co.uk/media/2012/oct/09/mobile-‐adverDsing-‐500-‐million-‐pounds-‐2012
“Connecting
with Travellers”
TrendGap
Channel
Vision
Data (any)
Insight
Business Traveller
going on a Business Trip
buys a ticket 7 Days or less before travel
52/41 = (127) index
Business Traveller
• 23% of all travellers, Frequent flyer (175)• male (123), 50 Years Old (139), Company Chairman (233)• Earns 90K / month (133)• Android smartphone / tablet user (NA)• Unaware of Airline Loyalty Program (109)
Behaviours (unique insight)
• Books travel & hotel < 7 days in advance (127)• Pays for tickets at airline ticket office (115), airport (117) airline call centre (111) and travel agent (115). • Checks-in at airline desk (106)• Buys advertised goods at airport (128), in-flight (113), later (117). Spends over 250 (191) on perfume, electronics, watches, sunglasses, toys, DVDs/Music• Research In-Flight magazine (114) and travel sites (132). • Plans next trip going to airport (129), in airport lounge (129) and on the flight (111)• Uses laptop on the plane (124)• Doesn’t value airline membership (67% of business)
Personality• Last minute business trip planner • Appreciates quality• Will use something if it makes his life easier• Wants respect and a personalised service. • Prefers human face-to-face contact• Impulse buyer, influenced by advertising• Likes to stay informed and up-to-date • Wants to maximise work time
Barriers• Does not trust social media recommendations• Does have time to do much research• Money not an issue but wants to know it is well spent
Output, Outcome, Impact
• Deal alerts <7 days before travel to increase Deal
Section Visits by 25% & Ad CPM by 25%• Improve mobile usage (095) and mobile boarding pass value (101) to reduce non-web and non-mobile bookings (all of above)• Faster mobile check-in (092) to reduce physical in-person check-ins (by all of above) and check-in time by 25%.• Product and service recommendations (200): business (1.35), food (1.06), hotel (1.03) and consumer products(1.04), entertainment (1.05) to improve Ancillary CPM by 200%• Personalised Ads: Automotive (1.31), Corporate (1.29), Conference (1.19), Electronics (1.10), Tourism (1.01) and in-flight retail choices (0.81) to increase online/mob CPM and conv (TBD).• Travel updates (0.94) and info to increase recommend to friend (092) and repeat booking purchase (101)• In-flight wifi to increase laptop usage (124)•(numbers) represent indexed data analyDcs
Deal Alerts <7 Days in advance of travel
Data
Empathy
Books travel & hotel <7 Days in advance (127)
Deal Alerts <7 Days in advance of travel
Increase Ancillary CPM rate by 25%
Does not have time to do research
Increase PVs by 25%
Last minute planner
OUTPUT
OUTCOME
IMPACTBarriers
Personality
Behaviour
(numbers) represent indexed GPS data
Success
OutputSoftware we build. The materials we produce.
OutcomeThe immediate change due to delivering output.
ImpactThe larger change we see over time
Outcome
Impact
Output
Assumptions
We believe by Output, we will Outcome* leading to Impact*
We believe by creating deal alerts, we will increase visits to deals section by 25% leading to an increase in Ad CPM by 25%
*QualitaDve or quanDtaDve measurement
Outcome
Impact
Output
What do we do first?
Value (relative)
EASY
HIGH VALUE
DIFFICULT
Quick wins Differentiator
Parity / building blocks Bin
LOW VALUE
Deal Alerts
Home screen advert
Basic geo-location
*if we had to build the capability ourselves. Therefore... Get client to build and enable
Mobile CMS
In-flight wifi*iOS Push Alerts
ROI:Customer value x PersonasBusiness value ABusiness value BEase of implementationDependency level
WireframesDeal Alerts <7 Days in advance of travel
My Flights
ALERT:Manila to Cebu
Book A Flight
ALERT / ADVERT
Flight Status
CP100 - OK
Check-in
Available (1)
Destination Guides CEB Club
Commercial Partners About Cebu Pacific
Sign-inRegister
Shake Iphone to view more Special Flight Offers
MANILA TO
BORACAYOne way from
P780
DISRUPTION: Manila to Cebu delayed
Flight notifications
Ancillary service notifications
OFF
Connect with Facebook
Disconnect Twitter
Connect with LinkedIn
Settings
Share this app
Contact us
Edit your personal details
CEB Club profile
Get in touch
Social Networks
Push notifications
ON
CEBU PACIFIC: You've just arrived. Book a taxi
and save 15% today
CEBU PACIFIC: Discounted fares to Boracay this
week starting from PHP 780
SPECIAL OFFERS
Discount Airtime Tariffs
SPECIAL OFFERS
ALERT / ADVERT
MANILA TO
BORACAY
One way from
P780.....Terms & ConditionsAdditional info on Special offers
BOOK NOW
Done
Outcome
Set explicit notifications (time, location, price etc.)
Receive well timed push notification
See on page alerts on key pages* (e.g. HP)
Click to view target page (Destination Page + AD)
*Assume persistent signed-‐in state
Impact
Output
Output
Output
What more can we measure?
Passenger info from Airline
Origina'onDes'na'onGenderDeparture dateReturn dateLanguageCabin classAgeAirline loyalty levelTicket type (open or return)Na'onality
Implicit
Click path analysisPage view / PreviewTime on pageShared on FB and TwiHerEmailGeo-‐locate / Check-‐inTime of day
Implicit but heavy AI, Machine Learning*
Sentence analysisMood analysisOther behavioural paHern analysis
Explicit
WizardsQues'onnairesBookingSearchSaveSeSngs / PreferencesLikes
Good Start Differentiation
Track it
What about making money?
Success can be Everywhere*
Triggers*
Holistic
AARRR!**
*BJ Fogg, Behavioral Analysis: h'p://www.behaviormodel.org/** Dave McClure: Pirate Metrics: h'p://www.slideshare.net/dmc500hats/startup-‐metrics-‐for-‐pirates-‐long-‐version
Don’t forget to experiment
*In App customer survey feedback form
Validation
• Guess
• Consequence
• Experiment
Cornell U, 1964
• Assumption
• Validate
• MVP
September 2011
Richard Fenyman Eric Ries
If it disagrees with experiment, it’s WRONG!
In Conclusion
• Problem / Goal• Opportunity• Data• Success= Outputs, Outcomes, Impact• Assumptions• Empathy• Value Map• Wireframes• Track & Capture• Make Money• Experiment
Metrics Driven DesignHow to define success and deliver more impact
Eewei Chen@Ultraman
http://101DesignIngredients.com