'ruaql asn cl1 .^doq liuru:eei lsnl e:rt oq-tt eldoed ro-J [6:r:adse(1dde or prtrl pue 'q8nogr 'xaldruoo uerjo ":-r.daqa'sosodrnd pougap-l{o-^tU 01 peroftl or,deqr osnelaq InJlln{ ore spoqlarlr pezrpnsd5 'ilolsrtreprlrunrl trnpord t aprn8 ot 1Jo.&!arue:IJ s.Jsllod ltodxa PInoA\ arro oN

,"soltrado tury 1r tl:rrql\ ur lxeluof rlrtsnpur erll urtldxe Pue sqrlrsapurl sel-ro3 aAU s(-reilod 'sseursllq e 3o ql,uor8 aql 3f,utug ol poqlaru lsaqaqt Sursooqc ro alrd :tp:rlred s 1t irnpo-rd e LIfunEI o1 -raqlaq,u Surpr:r

-ap roJ InJdleq Lra,r l.rrcre f,eqr rnq-a8etuelpe aarltlodruo: Surlt8r1se,turuar{Ar e^ulnpord e-rt sldc:uor s.ratJod IJer{lrW'suolltnlIsJo sad{l rgr:edspu?tsJepun dlari of parclrpr Jft ull[]Jo [\r'sloo] arululrlutlb pue'se:lrl:erdput sesseroJd 'syo,tetut,r3'ser-ro*{t u-{\o str stq eurJdosrp ss3ursnq qJtE

'uolltnlrs Iltro o aLIt ot uorltlst LIt u?orlr (eqr rtq.u PUt uorlenlrs t3o sl:edst luelroduir Su4rutlsrapun put 8ur,'furiuopr-Jo s[t,{r peau oro]3JJI{lnof 'rse] r aziltue oJ'rloq.tr :ql ot sutd eqr 3o sdrqsuontia-t ot{] {piusot pue :strJtd luenlBsuol slr olur dn 8rfiqtsruos :{e3]q ol :suollrulJoP polElardlasc'r1:r oA\t srq srsdpur 'qsq8rrg u1 .,'3ur,r,1ossrp e,,'Surueaul p-ror!{ la;lf) e

rrro-r+ pe^r-Iop sr srsf,lotty 'uon?ntrs B lnoqP srrolsl-llruol .tttrp clleq ol srs,(1tue

s.(oldrua ttqt 8urur?ol paprtr?-;1as -roJ rlrat E -1!trsuffit! sI porllsul es€l srlJ

3NI(INVJ.SUIICNN UOd TNIOiI CNIIUVTS

'asrc t Surzditue roJ poqleru t ol lunorut luop lieqr lng'srol'rg lenl-x3luol aseql I[tJO Jsn a>li]rll ppoqs no 'sanssr lue]rodrui ol uouuslle IItruel suoDsenb .(pnrs prre 'asrl .ttau t lnoqr 3ur1urq1 f,oJ uoqtpunoJ E apr^-ord suorssn:srp astl ts€d 'poultel a.ttq nof, serrlrtrr{ lseq put sioor p:p.!-rue aqt 1fo^\ ol rnd o1 sailiunlroddo pue selnpour 3tllJo utd ert ]?ql scst:leqt ur saluaq] aql r3tunolue ol t:adxs ut:l no 'suorltnlls essql ro-1 aletrd-o-rdrjr sa:rtou-rd put 'serroaqt 'stdecuoo 'uaqo 'pue suorlrnlrs 3o sodXr urer-:ar .{q perrgop seurelll Jo salnpotu tu3f,e-}Ip otrll poprllp uoryo JJll saslno'l'.ra8rueru Suunl:r:3nur:ru ;o tsr.3altf,ls t lou ';ele1:tur e s? Iu(il o1 rtodseJrnbar asto 8urla1.rtru V'esrlol t ur polpllls dllensn aJt sose]'srs,(pur aqt.\\orfeu Pur lrru{ dlaq reqr sJoltr?.} Fnlxsluof oruos SulzruSo:a-r Lq lrttS

.tr -ro3 )lurueottt LJre lrnrlsuoo nod op A\oH lleslr rirrldxa ol sasryel l"ql lxal , ,, .trt I

ASY) V flZA'TVNV OT /VTOH

E UAIdVHf,

I; il7 iltri-i.JS '*afinH tri3"lpti.d 'i*Jd 'Alisj*.{!irtl *i

2 ANALYSIS

r.3. { I

This book teaches an approach to cases that complements business con-cepts and theories. Its purpose is to provide a starting point for analysis thataids tire use of theories and frameworks and quxntitative formulas, all ofr,vhich are indispensable for reaching conclusions about a case and building an

areiument for those conclusions.The case situation approach identifies fea-tures of a case that can be helpful to its anaiy,sis and encourages active reading.

THINKING, NOT READING, IS KEY

Students new to the case methocl usually believe the rnost reliable way tc)

understand a case is to r-ead it from start to finish and thern reread it irs manvtimes as necessary. (That's why many business school students think speed-

reading courses can help them.) They rush into a case, highlighter in hand,reading as if the case were a textbook chapter. For case analysis 1'ou need toknow when to read fast and when to read slowly.You should also spendmore time thinking about a case than reading it.

Wiren -vou begin work on a new case, you dont kno'rv u,hat to look for.That is the major dilenma that corfionts everyone who reads a case. In an

active approach to a case, you start thinking b{ore you read the case. Afld as

you start reading it, you ask questiorx about the content. Then you seek

answers in the case itself.As you find partial or full answers, you think abouthow they relate to each other and to the big picture ofthe case.You dontmake knor,vledge by reading. Reading is never the primary lrcsource of case

analysis, Reading is simply an instrument directed by the thought process

that makes meaning lrom the text.

TYPES OF CASE SITUATIONS

Four rypes ofsituations occur repeatedly in cases:

. Problems

. Decisions

. Evaluations

. Rules

People sometimes react indignantly ro this classification.They insist thatthere are a multitude of situations portrayed in cases, and it's misleading tosay they're reducible to four.The four are not the oflll silLrrli.rs lound incases, but many case situations do belong in one of the four categories, andwhen they do, an arvareness of which one can help organize analysis.Thisapproach isnt the only correct way-ir is one way.Tiy it and see if it helps.

f*h F*ng *r ri,=la

r

:{

HOW TO ANALYZE A CASE

Feel free to integmte pieces of it with your owr way of dealing with cases.

The greatest value of the case situation approach may be that it causes youto think about how you think about case studies.

Problems

The wordproblemhas many meanings.The meaning can be vague, referringto something thatt difficult or troubling.The definitiot of problem as a case

situation, howeveq is quite specific. It is a situation in which (1) there is asignificant outcome or performance, and (2) there is no explicit explanation

of the outcome or performance.To put it simply, a problem is a situation inwhich something important has happened, but we dont know why it did.

Cases provide marry examples of problems defined this way. In one, a

well-trained, well-intentioned manager has tried to introduce a wofihwhilechange in the sales strategy of an organization-a change supported by adetailed, data-driven analysis everyone admits is a breakthrough-and has

failed to get any of the sales staff to go along- In another, an accountingmanager of a manufacturer notices that two good retail customers suddenly

have accounts payable that are large and overdue enough to be worrisome.He has no idea why the two firms would fall so far behind in their payments.

Both of these cases describe situations that involve negative outcomes.The causes of these sorts of outcomes are lmportant to know for a p al

ritaso&: the knowledge can help implgve the situation. The change effortmay be self-destructive because it ha\tr{veaknesses that are not apparent, orthe manager may be good at many things but is a poor change agent.Themanufacturer's retail customers may have large accounB palable because

they have sloppy internal controls-or they may both be on the verge ofbar*ruptcy.These possibilities illus*ate why accurate carxal analysis is vital.A conceptually flawed clnnge is addressed very differently &om an individ-ual who isnt well suited to lead change. If both situations exist, the correc-tive action is that much more complex. Retail operations that need to clean

up their accounting processes might require the manufacturer to engage innegotiations over a period of time, but two firms with bad debts that mightgo bankrupt require the supplier's immediate attenlion.

Success can also be a problem in the special meaning used here.Take thecase of a company that specializes in outdoor adverrising. It operates inthree different market segments, but the case doesnt tell you which is themost profitable, much less why.Another case describes the development of a

country over a period ofthirty years or so; after severe political and social up*heaval, the country slowly recovers and exceeds the performance of mostcountries in the region. But the case doesn? state how much more successfi.rl

trr*i ivlarcia Riih*n, $pring5.',iifil::*

4

4 ANALYSIS

the country has been relative to its neighbors, and while it provides a greatde:rl of data, both economic and demographic, it doesnt erllmerare the rea-sons for the country's revival.

Problem analysis begins rvith a definirion of the prol'ienr. That seems

obvious, yet many cases don't state a problem. So first, you need to realize a

problem exists and then define it for yourself. Next, you .,vork out an expla-nation ol the problem by linking the outcome or performance to its rootcauses-this is the main rvork of problem analysis. To carry it out, you'l1need relevant tools, the specialized urethods ofbusiness disciplines such as

organizational behavior or opera tions mana genlent.

Decisions

Manv cases are organized around an explicit decision. The second para-graph of "General Motors: Packard Electric l)ivision" (r-eproduced in thisbook) begins with this sentence: " Ihe Product, Process, and Reliabiliq,(PPR) committee, which had tire final responsibility for the new productdevelopment process, had asked [Davidl Schramm for his an:rJysis and rec-ommenrlation as to wirether Packard Electric should cornmi.t to the RIMgrommet fr>r a 1992 modei year car." Like many cases, this one complicatesthat decision irnmediately: Schramm mrxt make up his mind rn ithin a rneek,

and the product developnrent people and manufacturine disagree over whichway to go.

The existence of an explicit decision is an irnportant distinction,becausenearly a1i business cases involve decisions. In marrv of these cases, howe.*er,

the decisions are implicit and dependent on another situarion. Let's take a

case described earlier that involves a problem: the outdoor adverrising com-pany. The case implies a decision: What is the best str:rtegry the companvshould purslle it the future? This decision can only be made after the com-pany's current strategy and how rvell it works are analyzed.

The clecisions featured in cases vary greatly in scope, consequcrrce, andavailable data.An executive must decide whetirer to launch a product, movea plant, pursue a merger, or provide financing for a plamed expansion-orthe presiderrt of a country must decide whether to sign a controversial tradeagreement. Regardless of the dimensions of a decision, analyzing it requiresthe follorving:

. Decision options

. I)ecision criteria

. Relevant evidence

-,n.- t

.-$t€os':

HOW TO A}'IALYZE A CASE

identifdng decision optiorfs is ofren easy because rhe case tells you u,'hatthey are.As soon as you encoirt ter a stared decision, you should look for a

statement of the alternatives.\If they arent stated, then the first goal ofanalysis is to come up u,ith plJuslb.ie decision options.

The most important part of a decision analysis is determining the crite-ria.A rational decision cant be made u,ithout appropriare criteria.A deci-sion case isnt likely to state criteria-they have to be deril'ed throughcarefLl study of the specifics of the case, rt ith the help of specialized methocls.The critcria are used to develop evidence to complete a decision analysis.The soal is to detenrrine the decision thar creares the best fit benveen theavailable evidence and the criteria. In the General Motors case, a possibledccision criteria is value to the customer.The reader needs to find evidenceinclicating which option delivers the greatest value to rhe customer. (Thatdoesn't settle the nlatter, though, because there are other criteria.)

One other characteristic of decision analysis deserves mention here.Ther-e is no objectively correct decision.The standard for a good decision isthe one that creates more benefits than the alternatives and has fewer or less

severe dor,vnsides.

Evaluations

Evaluirtions express a judggnent about the rvorth,value, or efrectiveness of a

perfonnance, act, or oufcorne.The nnir of analysis of an evaluation can bean individual, a group, a department, an entire organization, a country, or a

global region.An annuai per{orm:rnce evaluation of an en-rployee is a real-world example. So is a nerv CEC) evaluating the perfonnance of the com-pany she is nor.r, heading.An evaluation can also irwoive the assessrnent ofan act, such as a decision that has already been taken. Herc is an exarnple:

Frow the perspecthr of rurrcrtt ELi mernbcrs, do you agree with their dccisiou

to enlar{e the Union by ten new rnembeys?

Finally, an outcome can be the subject of an assessment.The comperidveposition of a companv, fbr instance, is the outcorne of nurnerous decisionsand peformances as well as conlingencies such as macroeconomic conditiorx.

Like decision analy.sis, evaluation requires appropriate criteria.-Withoutthem, there are no standards for assessing worth, value, or eflectiveness.As indecision analysis, evaluative criteria are inferred from the particulars ofa sit-uation with irelp front specializecl methods. Er,'aiuating a company's finan-cial performance over a five-vear period can be undertaken with a long iistof financial {brmulas, but the circumstances portrayed in the case corne iotoplay as well.The numbers may shorv that a company has a steadily declining

ieel h'; L *h r: 5iOEi?01 nrv*rsit\r, trrni. l'4ercia H n:rrl

6 ANALYSIS

pertormance over the period, but it stiil may be doing well because thenational econorny is slunrping and the conrpany is actually doing better:

than its competitor-s.An orrerall evaluarion expr-esses the best fit betrvec'n the evidence and the

cr:itelia. In the example just given, measured against purely financial crite-ria, the company is doing poorly.Yet, the cvidence pertaining to nracroeco-

nonric ancl competitive criteria alters the evaluation: in a tough market, the

company is actualiy per{orn:ring better than its peers.

Another requirelnent of eval.ration is that it include both positive and

negative sides. A leader has strengths and weaknesses, and both are irtcludedin an accurate evaiuation. Morcover, there may be aspects of the leader's

performance that are ambiguous-he has delegated power rvidely; but it is

too earl)i to tell whetirer the managcrs below him can hanclle the power.

Ancl this individualt performance as a ieader could be substantially affected

by factors outside his control-corporate headcy-rarters has interr,'enecl irr

his prornotion decisions and insisted that certain favorites be elevatetl even

though thev aren't the best-qualified candidates.

Rules

ilarrtitative nrethods\ can plovide critical information abor,it business sittt-ailor-i-i:For exmfE. srv rhcrc is a need to cor)lparc the raltre of a compan5'

when a specific condition exists---a partnership with another company-and when it doesnt exist.The way to calculate future cash values-one thatexperts and experience support as reasonably accurate-is net present

value.An NPV calculation is done according to a forrnula. Mathematically,there is a right way to perform the calculation; any other way provides an

inaccurate result.For rules analysis, you need to know:

The type of information needed in a situation

The appropriate rule to furnish that information

The corect way to apply the rule

The data necessary to execute the rule

Rules anaiysis e-xists in virtualll. every area of business.A breakeven cal-culation is a mle used in marketing. In manufacturing. quantitative nreth-ods are used lor process anallrsis, and accounting and finance consistprirnarily of mles.The scope of rules is very narrow. For the most part, theyare useful on.ly in specific sets of circumstances, brit in those cit:umstancesare very prociuctive.There is J con'ect .way to execute or pedorm the rule,

/

L

inr: r.:hi*h Fleni: *n ii1ilEi?fli3 G ,-1fi iv#l'sitv. Flri:f. fu4er*i,;* lt* b*n, Snrr n

HOWTOANALYZEACASE 7

and the output is of one type.A well-defined set of rules is needed to ana-lyze a company's liquidity.Those rules are the most usefirl in the situation,because they are designed to be. Each calculation specified by a rule has a

procedure that must be followed. If it isnt, the result is a meaningless num-ber. Each calculation yields a precise output ofa prescribed type (e.g., a per-centage less than or equal to zero).

Qualitatite methods are different &om rules.There are often many alter-native methods for obtaining the same or similar informarion. To analyzethe quality of leadership in an organization or its competitive strategy, thereare a large ru,lmber of methods to choose from. There is no prescribedmethod that provides correct idormation about competitive adr,zntage. Inmar:heting, two different methods can be applied to the same situation, canproduce very di$erent resulm, and can both be usefirl--or useless.A seconddifference between rules and qualitative methods is how they are executed.There is a correct way to execute a rule such as the forrnula for net presentvalue; there is no objectively couect way to execute quaiitative methods forarnly zing co mperition.

That is not to say that rules analysis lacks uncertainties and amtriguities.Any calcularion about the future involves unceftainty. This uncertainry isbuilt into formulas through assumptions, and assumptions involve judg-ment, not objective tnrth- Settling on a growth or inflation rate over a cer-tain period of time is speculativa The key is the reasoniog behind thechoice. Central bankers can be wrong atrout inflation and growth, and so

can the rest of us,Assumptions need to have a reasonable basis, but reason-able people can disagree about them. But note that the argument is absutassumptions, not about the rules themselves. (E*perm do argre about thefitness of rules and make changes to them, but a{ier they do, everyoae usesthe changed rutre aad executes it the same way.)

Somecimes, though, an idiosyncratic assumption has no material effecton the result of a calculation. In the earlier valuation example, you mightassume a growth rate tltat is too optimistic, but if the rate is the same for thecalculation with and without the partnership, it should have no efGct onthe comparison of the end values.

The results ofrules analysis frequendy provoke sharp differences ofopin-ion.What two people infer from the same numerical results can diverge.Economists are famous for looking at the same set of numbers and comingto vasdy different conclusions about thern, even though they all agree onthe formulas and data that have produced the numbers.The same is true incompanies. One executive can read financial numbers as conirmation thata stratogy is working, wh:ile another can read them as a warning that disas-ter looms. In short, numbers dont e4plain what they mean, and they dontmake decisions for you.

ra"IF

7

7

il 5l{i*,rlilina e!'ri*h f( t*ber':. S*ri

8 ANALYS1S

l{orvever, the interpretation of the output of ruies is distinct &om the

rules themselves. If the right rule is applied and correctly perlormed, and the

rule doesnt involve a controversial assumption (ike dre predicted gro\'vth rate

of GNP), everyone will come up with exactly the same result. If a qualita-

rive method relevrurt to a situatioll is applied to the sarne set of facts in a rvay

consistent with the generally understood meaning olits concepts, everyone

rvil1 not necessarily come up rvith the same result.That is the fundamental

di{1-erence between ruies, as defined here, and qualitative methods-

Rules arent pursued further in this book. Learning rules analysis means

learning a certain categorv of rules-valuation, for instance-and when

and how to use them.That iearning is the province of accountinEs, finance,

tax, and other areas that are intensetry rule governed. F{or.vever, it may be

helpful to remember that when rules depend upon assumptions, the values

chosen for them require an argument. Moreover, the informarion rules provide

has great importance tbr the analysis of problems, decisions, and evalua-

tions.Accounting rules can diagr<lse the financial health of atr organization.

Macroecononrics is inr,-aluable in evaluating a nation's developnleut strat-

egy. Financial rules are indispensable to a decision about whether to sell a

company at a gi',,en time and price. Rules are a large and important subset

of the specialized methods necessary to understand case situations.

CASE ANALYSIS AS A PROCESS

The way you analyze a case differs from the \,vay anvone else does.There is

a clitlerence, though, befween personal study habie and a process for ana-

iyzing a case"The latter involves more than habits and practices. It concerns

how 1,ou think about a case. The intention of this section is to suggest a

process that has helped case method students become more elficient and

productive.This process is designed for case discussion preparation, but it is

easily adapted to a process for u,riting a case essay. (Hon'ever, the way a case

is analvzed for ln cssay is nrore prescriptive,sint'c an essay nltrst havc certain

elements. Chapters 10 through 12 u,ill explain these elements.)The key to the process is active readirrg" Active reading is interrogative aod

purposeful.You ask questions about the case and seek answers. Questionsgive a purpose for reading; they direct and lbcus study on important aspects

of a situation.The rnoment you sense that yotr are reading without purpose,

stop and regroup. It may be a good time to step away and strerch, do some

-voga, or walk- Active reading is also iteratite, meaning you make multiplepasses through a case.W-ith each iteration, the purpose of reading changes:

you are iooking for new information or looking at old jgformati,qp in a

new way Three concepts contribute to acti\re .""airrg, ,$ol, , pirrt ofvieui and a-hypothesis.

o

il]; *i$fft!t]s i]' li,'rli *olden Gat !qLihen Sf

HOW']]O ANALYZE A CASE

Goal of Analysis

At first it rnay seem obvious.lil/hat other goal can there be lbr analyzing acase than to understand ir? The probleni is that "unclerstanding,,is toovague.Another way to think about the goal is, How do you know when toconclude the st*dy of a case? This is an ir.nportant question. Il you tlonthave a concrete linrit,you can drili arong for hours,,ruch ofit taken up bydistracti.n and u.directed elrort. Here is a nlore concrete goal: vou arelamiliar with the in{brmation in the case, you ha'u,e .orr. ,o , c.nclusionabout the main issue, you have evidence showing why -vour concrusion isreasonable, and you have thought about other possible conclusions and ."vhvyours is preferable to them.

This substantive goal can be combined rvith a time limit. Allocate a seramount of time-two hours, for example--for each case.At the end of theperiod, stop and settle for whatever you knor,v about rhe case.This is a verygood rvay to put constructivp prcssurc on yourself to nrake the most of thetime. qotiel.e

Point of View

T. anchor analvsis, take advantage of rvhat's already i, the case.Adopt thepoint of view of the protagonist-the,rain character. put yourself in hershoes. Her dilemma sho,ld be your dilemma. lf it'.s a <iecision, set a recom-mended decision as your goal. whe, you adopr the persona of the mai'character, doni assume that -vou're dealing with a car<rboard cutout, a clra-maric veneer. consider rhe character's stren*hs, responsibilities, and blindspots. By all means, too, be sensitive to the dilemmas characters find them-selves in. often, a good question ro ask yourself is,w1ry is the person in thisdilemura?

Hypothesis

one of the most useful constructs 1or resoiving the protagonistt dilemma is ahypothesis.Ahypothesis is "a renradve expranarion thar acc.unls fbr a set of llctsand can be tested by further investipption."2 It is indispensable to scie,ce a'rlt. a,y fact-based anal).tic activiry- in r.vhich multiple concl'sio.s are possible.

A hypothesis olrers rhe advantage of a concrere statenrent you can testagai,st case evidence- Say that the protagonist of a case must evaiuate anindividual she has hired-a rising star-, but also a person who arienares ma'ypeopie inside the fir,r a,d cuts some corners in his relentless pursuit ofnew business.The h,vpothesis is that dre new hire should receive a high rat ngdespite some flarvs in his performance. To test it, 1r<ru ll har.,e to ,leo-elop a

*i"cii: *,;D*lt. S:aci*ii

1O ANALYSIS

strong argument, based on relevant criteria, facts, and inGrences, that backs

a positive evaluadon but also recognizes poor peformance on other criteria.Cases dont allorv just any hypothesis.The available evidence in the case

sets the rational linit on the range of hypotheses.A hypothesis that canl be

argued from evidence in the case is simpil, an unsubstantiated opinion.Flor,vever, there is a range of possible hypotheses allout every case. A con-trarian's position-one that opposes what seem to be safer h1'potheses and

can be argued frorn evidence-can have a galvanizing efrect in a discussion,

forcing everyone to look at the evidence from an entirely nern' angle orconsider evidence no one else has noticed.

DESCRIPTION OF PROCESS

lfhe rest of this chapter oudines a process for r,vorking on cases.The process

has five phases:

1. Situation

2. Questions

3. Hl,pothesis

4- Proofand action

5. Alternatives

The process is ureant to be flexible and adaptable. Experiment with it.using the cases in this book. Many MI3A students dont give much thoughtto their case-siudy approach, not because it is unimportant but because

rhey dont see anything tangibie to think about. Lrltinlateiy, the value of tireprocess described below depends on rvhether it prompts you to thirr& aboutYOUr Ol,Vn prOCeSS.

1. Situation (5 minutes)

The most dilEcult part of a case analysis seems to be the beginning.Youharre to bridge the gap between no knowledge about the case and knou4*edge sufiicient to form a hypothesis"That gap can look very wide as y'ou

begin reading a case thick r,vith detail; it can seem to be all parts and nowhole. Earlier in the chapter, I srressed that it is hard to find somethingrvhen you don't know vu'hat you're looki-ng for. To get started, you can

structurkafraiysis r,vith a series of questrjons.The process I advocate is under-standing$e blg p.ictr-Lqqf rst and thenYlling it in with detaiis. Start by ask-

ing this question; What is the situation?

f*i Meri s,; lll i 'r

HOW TO ANATYZE A CASE 1A

- Usually reading the first and last secf,iorx of the case is str{Ticient to iden*t& the situation.Decisions and evaruatiorx tend to u" ro,.i *rr" u"g;*"g

Problenu are harder to recogrrizg and more details about ia.rrti6rrrrg ;;* ,r"provided in chapter 5-A chamcteristic of a protrrem case is the atrsence ofany actionabre statement made by or abort tie protagonist. often, the maincharacter is reflecting on a situation and wonderiog irhr, to do-Reading the first and last secfioru of the case can often pro,ide far moreinformation than just the type of situation. In decision .rio, th.r" *.rio*

may speci$ the decision options.That is true of the case..General Motors:Packard Eleetric Division-" If you doat find options at the begin'ing orend of a case, you should scan other secrions.The opening o, .iai1g of *grobj"* case may present a qarrial or complete dooriptio.ior*r" f*Bt"*.In all types of cases, the initial and final secions frequefftry *ap.o, I o*io,or conflict important to the analysis. In ..Generzl Moton,,,ttie firrt sectionidentifies the decision and a conflict between two functional g*rfr. rlr"turc sides of the conflict" with the protagonist in the middre,

""'rn # ,.rrr-ence points for analysis. Why do the product development p.opi* ,ostrongly supporr an innor'ative componenr that they're *iiri"g r. Lt"', r"r-midable risk? And r,'rry are the rnanufaeturing peopre just as adamaat thatthe company should not go forward with the i*por"o, in the short term?

. A*er readiag the opeaings aad closing sections, you should put the caseaside for a mo,'lent and consider *hrt y-lo have learned. Is the situation aproblem, decision, or e.r,aruation? Do you have any ideas about the causarframeworls or criteria that might fit the situatisn? Does it seem you,lr haveto cut through a rarge am.urlt of inform*tion i,, the case or inu*

^rrryinferences because the information is scarce? Are there any hints in the rwolectilns about causes, crj*erta, or eveil a prausible decision or evaluation?Ds the hints seem reliable orjust , *y t, th.o* you ofi

t*:* the si,uation ;ff:: tr;",:Jr*#,: a prob,em,a decision, or an evaluation-The *o*t i*plrt ri ortu"u" questions is:whatdo I nced to know abowt the situation?Here are questioas specific to each situation:

PROBLEM'Who

or what is the subject of the prot{em (e.g., a manager, a company,a country)? what is the probter.nJ Am I rring to account for a failure, aurr::r, or somerhing more ambiguoroi linaJt the dgoi6can"r;;;problem to the subject? Who is responsible for the p.IU"* q"rUfy O, i,

wniuad*d i: lai* ilrlivr:rs

12 ANALYSIS

tire protagonist) and what nright he need to know to do soneethingab<iut it?

DECISION'What are the decision options? Do any seem particularly strong orweak? What's at stake in the decision? What are the possible criteria?What rnight the rnost important criteria be for this kind of decision?Are any of the criteria explicitiy discussed in the case (case heading;can sometimes give good clues)?

EVALUATION'W'ho or what is being evaluated? Who's r-esponsible for the evaluation?What'.s at stake?What are the possibie criteria?'lVhat nright the most

important criteria be tor this sort of er,aluation? Are any of the criteriaexplicitly discussed in the case (case headings can sonletimes give vougood clues)?

You u,ont be able to answer these questions now.That will take furtherstudy.To make this firstpass through the textmore tarlleted,it's usetui to doa con.tent in.uen.tory.Its purpose is to locate information that might be used toanswer the questions about the situarion.

To pedorrn a fast irrventory scan the headings in the text. Reacl a little olthe sections, especially those that seem to have vaiuable inforuration. Exarn-ine the exhibits to get a sense of what they convev.You r.rdl learn some-thing about the case----:iometimes a great deal more than 1,ou might expecrt.

You'll also build a map of the usefirl content. Because cases often ar-ent lin-ear in their organization, this map is very important; pieces r:f informatioorelated to the salne issue will be found in different sections of the case andin the ex1-ribits.

lJse a pencil or pen to mark up the case. Mark high-value sections andcircle facts, numbers, and statemcnts of possible importance. Be sure to cap-ture any thoughts about the answers to your questions. antl record newquestions that corne to mjnd. Note r.vhat issues particular exhibits may illu-n'rinate, and what calculations might be performed later to yield relevantinformation.

3. Hypothesis (45 minutes)

Armed with a list of thingp you \4,ant to know about the situation and a

nrap of the content, you are ready for this qtrestion: What\ nry hypothesis?

This is the most iurportant phase of work on the case. Through closestudy of high value sections and exhibits, you narrow the possibiiities to the

*h P*nS {rn 5i*S12il13 iI* LJni're,"$

HOW TO ANALYZE A CASE 13

one that seems most plausible to you. If there are three alteraatives for adecision, test them, starting with the one you suspect has the most pnomise.Flere are some other suggestions for structuring your work at this point:

PROBLEM

' Make sure you know the problem that needs to be diagnosed. con-sider whether the characteristics of the problem,rrgg*i causes.

' Think about the framewotks that seem most appropriate to the situ-ation. Quickly review the specifics of the frameworks ifyou arentsure of them.

' Pursue the diagnosis by looking at case information through the lensof the cause you are most certain about.

. For each cause, make a separate pass through the case looking forevidence ofit.

' If the case has a lot of quantitative evidence, to what cause is it mostrelevant? If you dont have a,cause relevaat to the quaatitative evi-dence, formulate one-'work up as much relevant, high-value quanti-tative evidence as you can.

. In a case with a protagorrist, consider whether she is a potenrialcause. Ifyou think she is, work out how she contributes to theproblem.

DECISION. Review the criteria you have come up with so far.Which do you

have the most confidence in?

. Review the decision options. Do any seem especially strong orweak?

' Apply the criterion fhat seerru to identify the most evidence inthe case.

' Investigate the strongest decision option with the criterion you havethe most confidence in. or, if you're reasonably certainatrout whichis weakest, see ifyou can dismiss that option quickly.

' Ifthe case has a lot of quantitative evidence,which criterion is mostrelevant to it? If you don't have a criterion retreva.sf to the quantita-tive evidence, formulate one-work up as much relevant, hijh-valuequantitative evidence as you can-

bv t-ing chreh Fer-rg nn S,ICB LTAIO Lin Sryrinr iiil

r

1.4 ANALYSIS

. lf there are conflicts about the decision between individuals orgroups, think about rvhy that is. Look at the decision from the pointofview ofeach ofthe parties to the conflict.

. If the protagonist is in a dillicult position in relation to the ilecision,considcr u'hy that is.

EVALUATION. Review the criteria you have come up with so far.Which do 1,ou

have the most confidence in?

. What are the telns of the evaluarion going to be (e.g., strengfhs,/

rveaknesses)? Do any stand out in the case (e.g., an obvior.rs strengthof an individual)?

. Do you already have a sense of the bottom-line evaluation youfavor? If .vou do, what are the reasons for the preference? Pursuethose reasons.

. Start by appiying the criterion that seenrs to idenrifi. the most evi-dence in the case.

. Investigate the most positive ratin5! or the most negative with thecriterion 1,ou have the most confidence in.

. If the case has a lot of quantitative evidence, which criterir:n is ruostrelerant to it? Ifyou dont have a criterion relevant to the quantita*tive evidence, formulate one. Work up as rnuch relevant, high-valuequantitative evidence as you can.

Trkirg notes helps you organize and remember information, but it serves

the equally important purpose of recording your tirought process.Withoutnote taking, you can too easily stray from active reading. Of course, notetaking can degenerate into transferring infbrmation in the case to a piece ofpaper or computer screen. Notes on a case don't simply record facts.Theycapture anything that might lead to answers to the questions you've asked.

It may sound trivial, l-rut I reconunend that studenrs try to contain the"highlighter habit."This study aid is weil adapted ro the lecture model ofIearning, but it can be a detriment to case study. Highlighting sentences is

satistying because it makes you Gel you're doing somerhing. In realiry rvhat

1.ou're doing is nurking senterlces to think about later, and that's a setup lbrpassive reading.You should be thinking about statemerts tlrc frst tinre youencounter them.That said, higlrlighters can be useful as a tool to difGrenti-ate related content: facts about one aspect of the case, for example, or te\.tand nurabers that belong to one cateplory of evidence.

Lifiiv*:'si

HOW TO ANALYZE A CASE 15

A pencil or pen is more conducive to active reading-to write downquestions and make notes.W.hen you begin to gravitate toward a conclu-sion,stop work aad write it down.The fimction of a hypothesis is to giveyou a position to try out, ltot a fi.nal conclusion, so listea carefully to yourintuition.

Ifyou have time, put the case away after t}is iteration. Even a short breakcan be usefi;l.There is scientific evidence that our subconsciorx min* aremuch better at dealing with complexity than our conscious minds.Turningyorrr attention to sornetling else allorars tlut subcorxcious capacity to workon the infornr.ation you have collected.

4. Proof and Action (4A minutes)

A hypothesis drives a diEerent approach to the case-You want to provesomething, not look for something to prove.Ask these questiorr.: What eli-dence ilo I have that supports the lrypothesis? What additional eyidewe d.o I need?

Look at the information you've compiled and ideatify evidence sup-porting the hypcrlresis.Yor:r firsr priority should be to add ro the evidenceyou have.What is the strongest evidence? Can you add more to it?

Now assess where evidence is missing.'Where will you 6nd more--or isthere arqz evidence in the case?Think about any factors you rnay have over*looked such as a cause, criterion, or evaluative category.

Go track into the case, with the single purpose of briaging out rnore evi-dence that aligns wrth your hypothesis.You dont have ro wo* from thefirst page to the last.You can go direcdy to the sections and exhibits youthink have what you need. Of course,you can work from beginning to endif that makes you rnore cornfortable. Just be zure fo stay focused on whatyou're tryring to prove.

Let's say that you're building an argunlent for a decision option and oneof the criteria is cost savings.You've noted some staterrrsnts that irnply yourdecision option will save money for the firm and circled numbers that youthought were relevant to savings. Collect those numbers now,and work outcalculations to estimate the total savings.You may then have one of thosegratifying moments of case sardy: from those scattered numbers that lookedso inconsequential when viewed individuaily, you've pulled together anestirnate that indicates a very large anaual savings-*and thatt just one partofyour argument.

Also give some thought to the actionable content ofyour position- FIowwould you implement the decision you're recornmeadiag?'What acfionsdoes your diaposis or evaluation call for? Tbink in practica\ real-wodd,aot ideal-qrodd terrns. Dont jlxt sketch out in your mind a broad

ih,vni* {l*d bv i-Ifi{j chi*i: F*ng CIn 5li}8/20'13. Gctr**n trA I*i'sitv" p ;[1';

16 ANALYSIS

approach to acfion. Think about tangible acrions and write thern down.Finally, give a bit of thought to the order of the actions.An acrion plan is aprogram in which acfions are taken at a, cerain dme for a reason. It isnt a

to-do list.

5. Abernatives (75 minutes)

It may seem paradodcal, but the last phase of analysis should be to questionyour own hypothesis: What is the greatest weakness of the hypothrsis? What isthe strongest alternatiue to it?

The inter*ion isn? to uadermine your hard work but to take a step backand look critically at the hlpothesis and the evidence. Every position has aweakness, and you should be the one who recognizes it, not the professoror your peen. Here are some ways to think critically about your work:

PROBLEM

Can the problem be defined difterendy?Would that make a differenceto the diagnosis? Are there any holes in the diagnosis---could there becauses rnisoing?lVhat's the weakest parr of the diagnosis? Could anentirely different diagnosis be made?'What would it look like?

DECISIONWhat's the biggest downside of the recommended decision? Howwould you manage the downside?Whatt the strongest evidence againstthe recommesdafion? How.urculd a case for the major akernative look?

EVALUATIONHave you been objective and thorough about the evaluative findingsthat oppose your overall assessmerrt? Think hour a different overallevaluation roight be proved. Have you accounted for factors that thesubject of the evaluation couldn't control?

..BUT WHAT IF MY HYPOTHESIS IS WRONG?"

Students have askcd me that qilestion rnany tirlles,A hlrpothesis isnt wrong;a hypothesis fails when you can't make a credible argument rbr it from caseevidence. If vou find 1,ourse1f in that situation-and you ni1l sooner orlater-first make sure the difirculty iies r.vith the hl,pothesis and not withyour evidence gathering.You may have overiooked important infbrrnatjonor not used speciaiized tools efGctively. If you're certaiir the evidence isn'rthere, lace r"rp to it but r-ealize that the work you've already done isnt u,astecl.

!]r: tr*iosd*d by [-ing chi*h F*ni; sn 5/,1*/2i]13. Gr:irier: ti,. Frot. I'"larcia *:-rben

HOWTOANALYZEACASE L7

You now have a good grasp ofthe case and probably have a good sense of

what the evidence is and where it is'Your work with a new hypothesis is

therefore likely to move along quickly'

Another way of l."ki"; ;4" f'* of being wroag is to ask'yourself

what the alternative ir. i f i'" not heard of a method of case analysis that

never leads to dubious conclusions' In fact' makiog analytic mistakes is

iroal*rbl". Through mistakes, we learn more about the thought proce6$

called case analysis- And a shaky analysis can sometimes be a qrmptom of

,l*t trti"g, *t lct it also an invaluable learaing experience'

NOTES

1. Michael E.Por tet,Competititrc Sildtegy:TahniquesfuAnalyzinglndusties antl Competi'

ron {NerrrYork:The Free Press, tr980)' r,-,^- E^+^6. rr^,,.,htnn M2. The Amerkan ,*'*i'' i'tW Diaio*ory'third edition @ostoo: Houghton Miflin

ComganY 1993).

Feirg r:n 5108i:*'i3. G*iel*n Sate li :lver*ity, 13l*f. Ma;"c:a Rli*err, S5:rlng ?CI111

llil'.*