How to analyze a case

How to Analyze a Case TheCase StudyHandbook: How to Read, Discuss, and Write Persuasively About Cases By Do NoHarvardBusiness School Press Boston, Massachusetts

This chapter was originally published as chapter 3 of The Case Study Handbook: How to Read, Discuss, and Write Persuasively About Cases, copyright 2007 Harvard Business School Publishing Corporation. No part of this publication may be reproduced, stored in or introduced into a retrieval system, or transmitted, in any form, or by any means (electronic, mechanical, photocopying, recording, or otherwise), without the prior permission of the publisher. Requests for permission should be directed to[email protected]harvard. edu, or mailed to Permissions, Harvard Business School Publishing, 60 Harvard Way, Boston, Massachusetts 02163.

The case method is heuristic—a term for self-guided learning that employs analysis to help draw conclusions about a situation.

Analysis is derived from a Greek word meaning, “ a dissolving. ” In English, analysis has two closely related de? nitions: to break something up into its constituent parts; and to study the relationships of the parts to the whole. To analyze a case, you therefore need ways of identifying and understanding important aspects of a situation and what they mean in relation to the overall situation. Each business discipline has its own theories, frameworks, processes and practices, and quantitative tools. All of them are adapted to help understand speci? c types of situations.

Michael Porter’s concepts are productive when investigating competitive advantage—but they aren’t very helpful for deciding whether to launch a product at a particular price or choosing the best method to ? nance the growth of a business. Porter’s ? ve forces can describe and explain the industry context in which a ? rm operates. 1 No one would expect Porter’s framework to guide a product launch decision. Specialized methods are fruitful because they’re tailored to ? t well-de? ned purposes. They’re often complex, though, and hard to apply, especially for people who are just learning how to use them. No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC STARTING POINT FOR UNDERSTANDING op yo A case is a text that refuses to explain itself. How do you construct a meaning for it? Start by recognizing some contextual factors that help limit and narrow the analysis. Cases are usually studied in a course. A marketing case requires you to think as a marketer, not a strategist or manufacturing manager. Courses are often divided into different modules or themes de? ed by certain types of situations and, often, concepts, theories, and practices appropriate for these situations. You can expect to encounter the themes in the cases that are part of the modules and opportunities to put to work the analytical tools and best practices you have learned. Past case discussions provide a foundation for thinking about a new case, and study questions can call attention to important issues. You should make use of all these contextual factors, but they don’t amount to a method for analyzing a case. rP os t 2 ANALYSIS This book teaches an approach to cases that complements business concepts and theories.

Its purpose is to provide a starting point for analysis that aids the use of theories and frameworks and quantitative formulas, all of which are indispensable for reaching conclusions about a case and building an argument for those conclusions. The case situation approach identi? es features of a case that can be helpful to its analysis and encourages active reading. THINKING, NOT RE ADING, IS KEY tC • Problems Students new to the case method usually believe the most reliable way to understand a case is to read it from start to ? nish and then reread it as many times as necessary. That’s why many business school students think speedreading courses can help them. ) They rush into a case, highlighter in hand, reading as if the case were a textbook chapter. For case analysis you need to know when to read fast and when to read slowly. You should also spend more time thinking about a case than reading it. When you begin work on a new case, you don’t know what to look for. That is the major dilemma that confronts everyone who reads a case. In an active approach to a case, you start thinking before you read the case. And as you start reading it, you ask questions about the content.

Then you seek answers in the case itself. As you ? nd partial or full answers, you think about how they relate to each other and to the big picture of the case. You don’t make knowledge by reading. Reading is never the primary resource of case analysis. Reading is simply an instrument directed by the thought process that makes meaning from the text. Four types of situations occur repeatedly in cases: • Decisions • Evaluations • Rules Do No People sometimes react indignantly to this classi? cation. They insist that there are a multitude of situations portrayed in cases, and it’s misleading to say they’re reducible to four.

The four are not the only situations found in cases, but many case situations do belong in one of the four categories, and when they do, an awareness of which one can help organize analysis. This approach isn’t the only correct way—it is one way. Try it and see if it helps. This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. op yo TYPES OF C A SE SITUATIONS rP os t Feel free to integrate pieces of it with your own way of dealing with cases.

The greatest value of the case situation approach may be that it causes you to think about how you think about case studies. Problems Do The word problem has many meanings. The meaning can be vague, referring to something that’s dif? cult or troubling. The de? nition of problem as a case situation, however, is quite speci? c. It is a situation in which (1) there is a signi? cant outcome or performance, and (2) there is no explicit explanation of the outcome or performance. To put it simply, a problem is a situation in which something important has happened, but we don’t know why it did.

Cases provide many examples of problems de? ned this way. In one, a well-trained, well-intentioned manager has tried to introduce a worthwhile change in the sales strategy of an organization—a change supported by a detailed, data-driven analysis everyone admits is a breakthrough—and has failed to get any of the sales staff to go along. In another, an accounting manager 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 ? rms 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 important to know for a practical reason: the knowledge can help improve the situation. The change effort may be self-destructive because it has weaknesses that are not apparent, or the manager may be good at many things but is a poor change agent. The manufacturer’s retail customers may have large accounts payable because they have sloppy internal controls—or they may both be on the verge of bankruptcy. These possibilities illustrate why accurate causal analysis is vital.

A conceptually ? awed change is addressed very differently from an individual who isn’t well suited to lead change. If both situations exist, the corrective action is that much more complex. Retail operations that need to clean up their accounting processes might require the manufacturer to engage in negotiations over a period of time, but two ? rms with bad debts that might go bankrupt require the supplier’s immediate attention. Success can also be a problem in the special meaning used here. Take the case of a company that specializes in outdoor advertising.

It operates in three different market segments, but the case doesn’t tell you which is the most pro? table, much less why. Another case describes the development of a country over a period of thirty years or so; after severe political and social upheaval, the country slowly recovers and exceeds the performance of most countries in the region. But the case doesn’t state how much more successful No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. C op yo rP os t HOW TO ANALYZE A CASE 3 4 ANALYSIS the country has been relative to its neighbors, and while it provides a great deal of data, both economic and demographic, it doesn’t enumerate the reasons for the country’s revival. Problem analysis begins with a de? nition of the problem. That seems obvious, yet many cases don’t state a problem. So ? rst, you need to realize a problem exists and then de? ne it for yourself. Next, you work out an explanation of the problem by linking the outcome or performance to its root causes—this is the main work of problem analysis.

To carry it out, you’ll need relevant tools, the specialized methods of business disciplines such as organizational behavior or operations management. Do No Many cases are organized around an explicit decision. The second paragraph of “ General Motors: Packard Electric Division” (reproduced in this book) begins with this sentence: “ The Product, Process, and Reliability (PPR) committee, which had the ? nalresponsibilityfor the new product development process, had asked [David] Schramm for his analysis and recommendation as to whether Packard Electric should commit to the RIM grommet for a 1992 model year car. Like many cases, this one complicates that decision immediately: Schramm must make up his mind within a week, and the product development people and manufacturing disagree over which way to go. The existence of an explicit decision is an important distinction, because nearly all business cases involve decisions. In many of these cases, however, the decisions are implicit and dependent on another situation. Let’s take a case described earlier that involves a problem: the outdoor advertising company. The case implies a decision: What is the best strategy the company should pursue in the future?

This decision can only be made after the company’s current strategy and how well it works are analyzed. The decisions featured in cases vary greatly in scope, consequence, and available data. An executive must decide whether to launch a product, move a plant, pursue a merger, or provide ? nancing for a planned expansion—or the president of a country must decide whether to sign a controversial trade agreement. Regardless of the dimensions of a decision, analyzing it requires the following: • Decision options • Decision criteria • Relevant evidence

This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC op yo Decisions rP os t Do No From the perspective of current EU members, do you agree with their decision to enlarge the Union by ten new members? Finally, an outcome can be the subject of an assessment. The competitive position of a company, for instance, is the outcome of numerous decisions and performances as well as contingencies such as macroeconomic conditions. Like decision analysis, evaluation requires appropriate criteria.

Without them, there are no standards for assessing worth, value, or effectiveness. As in decision analysis, evaluative criteria are inferred from the particulars of a situation with help from specialized methods. Evaluating a company’s ? nancial performance over a ? ve-year period can be undertaken with a long list of ? nancial formulas, but the circumstances portrayed in the case come into play as well. The numbers may show that a company has a steadily declining This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.

[email protected]harvard. edu or 617. 783. 7860. tC Evaluations express a judgment about the worth, value, or effectiveness of a performance, act, or outcome. The unit of analysis of an evaluation can be an individual, a group, a department, an entire organization, a country, or a global region. An annual performance evaluation of an employee is a realworld example. So is a new CEO evaluating the performance of the company she is now heading. An evaluation can also involve the assessment of an act, such as a decision that has already been taken. Here is an example: op yo Evaluations

Identifying decision options is often easy because the case tells you what they are. As soon as you encounter a stated decision, you should look for a statement of the alternatives. If they aren’t stated, then the ? rst goal of analysis is to come up with plausible decision options. The most important part of a decision analysis is determining the criteria. A rational decision can’t be made without appropriate criteria. A decision case isn’t likely to state criteria—they have to be derived through careful study of the speci? cs of the case, with the help of specialized methods.

The criteria are used to develop evidence to complete a decision analysis. The goal is to determine the decision that creates the best ? t between the available evidence and the criteria. In the General Motors case, a possible decision criteria is value to the customer. The reader needs to ? nd evidence indicating which option delivers the greatest value to the customer. (That doesn’t settle the matter, though, because there are other criteria. ) One other characteristic of decision analysis deserves mention here. There is no objectively correct decision.

The standard for a good decision is the one that creates more bene? ts than the alternatives and has fewer or less severe downsides. rP os t HOW TO ANALYZE A CASE 5 6 ANALYSIS Do No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC Quantitative methods can provide critical information about business situations. For example, say there is a need to compare the value of a company when a speci? c condition exists—a partnership with another company— and when it doesn’t exist.

The way to calculate future cash values—one that experts and experience support as reasonably accurate—is net present value. An NPV calculation is done according to a formula. 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 correct way to apply the rule • The data necessary to execute the rule Rules analysis exists in virtually every area of business. A breakeven calculation is a rule used in marketing.

In manufacturing, quantitative methods are used for process analysis, and accounting and ? nance consist primarily of rules. The scope of rules is very narrow. For the most part, they are useful only in speci? c sets of circumstances, but in those circumstances are very productive. There is a correct way to execute or perform the rule, op yo Rules performance over the period, but it still may be doing well because the national economy is slumping and the company is actually doing better than its competitors. An overall evaluation expresses the best ? t between the evidence and the criteria.

In the example just given, measured against purely ? nancial criteria, the company is doing poorly. Yet, the evidence pertaining to macroeconomic and competitive criteria alters the evaluation: in a tough market, the company is actually performing better than its peers. Another requirement of evaluation is that it include both positive and negative sides. A leader has strengths and weaknesses, and both are included in an accurate evaluation. Moreover, there may be aspects of the leader’s performance that are ambiguous—he has delegated power widely, but it is too early to tell whether the managers below him can handle the power.

And this individual’s performance as a leader could be substantially affected by factors outside his control—corporate headquarters has intervened in his promotion decisions and insisted that certain favorites be elevated even though they aren’t the best-quali? ed candidates. rP os t Do and the output is of one type. A well-de? ned set of rules is needed to analyze a company’s liquidity. Those rules are the most useful in the situation, because they are designed to be. Each calculation speci? ed by a rule has a procedure that must be followed. If it isn’t, the result is a meaningless number.

Each calculation yields a precise output of a prescribed type (e. g. , a percentage less than or equal to zero). Qualitative methods are different from rules. There are often many alternative methods for obtaining the same or similar information. To analyze the quality ofleadershipin an organization or its competitive strategy, there are a large number of methods to choose from. There is no prescribed method that provides correct information about competitive advantage. In marketing, two different methods can be applied to the same situation, can produce very different results, and can both be useful—or useless.

A second difference between rules and qualitative methods is how they are executed. There is a correct way to execute a rule such as the formula for net present value; there is no objectively correct way to execute qualitative methods for analyzing competition. That is not to say that rules analysis lacks uncertainties and ambiguities. Any calculation about the future involves uncertainty. This uncertainty is built into formulas through assumptions, and assumptions involve judgment, not objective truth. Settling on a growth or in? ation rate over a certain period of time is speculative.

The key is the reasoning behind the choice. Central bankers can be wrong about in? ation and growth, and so can the rest of us. Assumptions need to have a reasonable basis, but reasonable people can disagree about them. But note that the argument is about assumptions, not about the rules themselves. (Experts do argue about the ? tness of rules and make changes to them, but after they do, everyone uses the changed rule and executes it the same way. ) Sometimes, though, an idiosyncratic assumption has no material effect on the result of a calculation.

In the earlier valuation example, you might assume a growth rate that is too optimistic, but if the rate is the same for the calculation with and without the partnership, it should have no effect on the comparison of the end values. The results of rules analysis frequently provoke sharp differences of opinion. What two people infer from the same numerical results can diverge. Economists are famous for looking at the same set of numbers and coming to vastly different conclusions about them, even though they all agree on the formulas and data that have produced the numbers.

The same is true in companies. One executive can read ? nancial numbers as con? rmation that a strategy is working, while another can read them as a warning that disaster looms. In short, numbers don’t explain what they mean, and they don’t make decisions for you. No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC op yo rP os t HOW TO ANALYZE A CASE 7 8 ANALYSIS Do The way you analyze a case differs from the way anyone else does.

There is a difference, though, between personal study habits and a process for analyzing a case. The latter involves more than habits and practices. It concerns how you think about a case. The intention of this section is to suggest a process that has helped case method students become more ef? cient and productive. This process is designed for case discussion preparation, but it is easily adapted to a process for writing a case essay. (However, the way a case is analyzed for an essay is more prescriptive, since an essay must have certain elements. Chapters 10 through 12 will explain these elements. The key to the process is active reading. Active reading is interrogative and purposeful. You ask questions about the case and seek answers. Questions give a purpose for reading; they direct and focus study on important aspects of a situation. The moment you sense that you are reading without purpose, stop and regroup. It may be a good time to step away and stretch, do some yoga, or walk. Active reading is also iterative, meaning you make multiple passes through a case. With each iteration, the purpose of reading changes: you are looking for new information or looking at old information in a new way.

Three concepts contribute to active reading: a goal, a point of view, and a hypothesis. No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC op yo However, the interpretation of the output of rules is distinct from the rules themselves. If the right rule is applied and correctly performed, and the rule doesn’t involve a controversial assumption (like the predicted growth rate of GNP), everyone will come up with exactly the same result.

If a qualitative method relevant to a situation is applied to the same set of facts in a way consistent with the generally understood meaning of its concepts, everyone will not necessarily come up with the same result. That is the fundamental difference between rules, as de? ned here, and qualitative methods. Rules aren’t pursued further in this book. Learning rules analysis means learning a certain category of rules—valuation, for instance—and when and how to use them. That learning is the province of accounting, ? nance, tax, and other areas that are intensely rule governed.

However, it may be helpful to remember that when rules depend upon assumptions, the values chosen for them require an argument. Moreover, the information rules provide has great importance for the analysis of problems, decisions, and evaluations. Accounting rules can diagnose the ? nancialhealthof an organization. Macroeconomics is invaluable in evaluating a nation’s development strategy. Financial rules are indispensable to a decision about whether to sell a company at a given time and price. Rules are a large and important subset of the specialized methods necessary to understand case situations.

C A SE ANALYSIS A S A PROCESS rP os t Goal of Analysis No Do One of the most useful constructs for resolving the protagonist’s dilemma is a hypothesis. A hypothesis is “ a tentative explanation that accounts for a set of facts and can be tested by further investigation. ” 2 It is indispensable toscienceand to any fact-based analytic activity in which multiple conclusions are possible. A hypothesis offers the advantage of a concrete statement you can test against case evidence. Say that the protagonist of a case must evaluate an individual she has hired—a rising star, but also a person who alienates many people inside the ? m and cuts some corners in his relentless pursuit of new business. The hypothesis is that the new hire should receive a high rating despite some ? aws in his performance. To test it, you’ll have to develop a This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC To anchor analysis, take advantage of what’s already in the case. Adopt the point of view of the protagonist—the main character. Put yourself in her shoes. Her dilemma should be your dilemma. If it’s a decision, set a recommended decision as your goal.

When you adopt the persona of the main character, don’t assume that you’re dealing with a cardboard cutout, a dramatic veneer. Consider the character’s strengths, responsibilities, and blind spots. By all means, too, be sensitive to the dilemmas characters ? nd themselves in. Often, a good question to ask yourself is, Why is the person in this dilemma? op yo Point of View Hypothesis At ? rst it may seem obvious. What other goal can there be for analyzing a case than to understand it? The problem is that “ understanding” is too vague. Another way to think about the goal is, How do you know when to conclude the study of a case?

This is an important question. If you don’t have a concrete limit, you can drift along for hours, much of it taken up by distraction and undirected effort. Here is a more concrete goal: you are familiar with the information in the case, you have come to a conclusion about the main issue, you have evidence showing why your conclusion is reasonable, and you have thought about other possible conclusions and why yours is preferable to them. This substantive goal can be combined with a time limit. Allocate a set amount of time—two hours, for example—for each case.

At the end of the period, stop and settle for whatever you know about the case. This is a very good way to put constructive pressure on yourself to make the most of the time. rP os t HOW TO ANALYZE A CASE 9 10 ANALYSIS strong argument, based on relevant criteria, facts, and inferences, that backs a positive evaluation but also recognizes poor performance on other criteria. Cases don’t allow just any hypothesis. The available evidence in the case sets the rational limit on the range of hypotheses. A hypothesis that can’t be argued from evidence in the case is simply an unsubstantiated opinion.

However, there is a range of possible hypotheses about every case. A contrarian’s position—one that opposes what seem to be safer hypotheses and can be argued from evidence—can have a galvanizing effect in a discussion, forcing everyone to look at the evidence from an entirely new angle or consider evidence no one else has noticed. The rest of this chapter outlines a process for working on cases. The process has ? ve phases: 1. Situation 2. Questions 3. Hypothesis 4. Proof and action 5. Alternatives No tC The process is meant to be ? exible and adaptable. Experiment with it, using the cases in this book.

Many MBA students don’t give much thought to their case-study approach, not because it is unimportant but because they don’t see anything tangible to think about. Ultimately, the value of the process described below depends on whether it prompts you to think about your own process. Do The most dif? cult part of a case analysis seems to be the beginning. You have to bridge the gap between no knowledge about the case and knowledge suf? cient to form a hypothesis. That gap can look very wide as you begin reading a case thick with detail; it can seem to be all parts and no whole. Earlier in the chapter, I stressed that it is hard to ? d something when you don’t know what you’re looking for. To get started, you can structure analysis with a series of questions. The process I advocate is understanding the big picture ? rst and then ? lling it in with details. Start by asking this question: What is the situation? This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. op yo 1. Situation (5 minutes) DESCRIPTION OF PROCESS rP os t Do No Knowing the situation allows you to ask questions pertinent to a problem, a decision, or an evaluation.

The most important of these questions is: What do I need to know about the situation? Here are questions speci? c to each situation: PROBLEM Who or what is the subject of the problem (e. g. , a manager, a company, a country)? What is the problem? Am I trying to account for afailure, a success, or something more ambiguous? What’s the signi? cance of the problem to the subject? Who is responsible for the problem (usually it is This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 83. 7860. tC Usually reading the ? rst and last sections of the case is suf? cient to identify the situation. Decisions and evaluations tend to be stated at the beginning. Problems are harder to recognize, and more details about identifying them are provided in chapter 5. A characteristic of a problem case is the absence of any actionable statement made by or about the protagonist. Often, the main character is re? ecting on a situation and wondering what to do. Reading the ? rst and last sections of the case can often provide far more information than just the type of situation.

In decision cases, these sections may specify the decision options. That is true of the case “ General Motors: Packard Electric Division. ” If you don’t ? nd options at the beginning or end of a case, you should scan other sections. The opening or ending of a problem case may present a partial or complete description of the problem. In all types of cases, the initial and ? nal sections frequently express a tension or con? ict important to the analysis. In “ General Motors,” the ? rst section identi? es the decision and a con? ict between two functional groups. The two sides of the con? ct, with the protagonist in the middle, can be reference points for analysis. Why do the product development people so strongly support an innovative component that they’re willing to take a formidable risk? And why are the manufacturing people just as adamant that the company should not go forward with the component in the short term? After reading the openings and closing sections, you should put the case aside for a moment and consider what you have learned. Is the situation a problem, decision, or evaluation? Do you have any ideas about the causal frameworks or criteria that might ? t the situation?

Does it seem you’ll have to cut through a large amount of information in the case or make many inferences because the information is scarce? Are there any hints in the two sections about causes, criteria, or even a plausible decision or evaluation? Do the hints seem reliable or just a way to throw you off? 2. Questions (15 minutes) op yo rP os t HOW TO ANALYZE A CASE 11 12 ANALYSIS the protagonist) and what might he need to know to do something about it? DECISION What are the decision options? Do any seem particularly strong or weak? What’s at stake in the decision? What are the possible criteria?

What might the most important criteria be for this kind of decision? Are any of the criteria explicitly discussed in the case (case headings can sometimes give good clues)? Who or what is being evaluated? Who’s responsible for the evaluation? What’s at stake? What are the possible criteria? What might the most important criteria be for this sort of evaluation? Are any of the criteria explicitly discussed in the case (case headings can sometimes give you good clues)? You won’t be able to answer these questions now. That will take further study. To make this ? rst pass through the text more targeted, it’s useful to do a content inventory.

Its purpose is to locate information that might be used to answer the questions about the situation. To perform a fast inventory, scan the headings in the text. Read a little of the sections, especially those that seem to have valuable information. Examine the exhibits to get a sense of what they convey. You will learn something about the case—sometimes a great deal more than you might expect. You’ll also build a map of the useful content. Because cases often aren’t linear in their organization, this map is very important; pieces of information related to the same issue will be found in different sections of the case and in the exhibits.

Use a pencil or pen to mark up the case. Mark high-value sections and circle facts, numbers, and statements of possible importance. Be sure to capture any thoughts about the answers to your questions, and record new questions that come to mind. Note what issues particular exhibits may illuminate, and what calculations might be performed later to yield relevant information. No tC Do Armed with a list of things you want to know about the situation and a map of the content, you are ready for this question: What’s my hypothesis? This is the most important phase of work on the case.

Through close study of high value sections and exhibits, you narrow the possibilities to the This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. op yo EVALUATION 3. Hypothesis (45 minutes) rP os t one that seems most plausible to you. If there are three alternatives for a decision, test them, starting with the one you suspect has the most promise. Here are some other suggestions for structuring your work at this point: PROBLEM • Make sure you know the problem that needs to be diagnosed.

Consider whether the characteristics of the problem suggest causes. • Pursue the diagnosis by looking at case information through the lens of the cause you are most certain about. • For each cause, make a separate pass through the case looking for evidence of it. • If the case has a lot of quantitative evidence, to what cause is it most relevant? If you don’t have a cause relevant to the quantitative evidence, formulate one. Work up as much relevant, high-value quantitative evidence as you can. • In a case with a protagonist, consider whether she is a potential cause.

If you think she is, work out how she contributes to the problem. DECISION • Review the criteria you have come up with so far. Which do you have the most con? dence in? • Review the decision options. Do any seem especially strong or weak? • Apply the criterion that seems to identify the most evidence in the case. • Investigate the strongest decision option with the criterion you have the most con? dence in. Or, if you’re reasonably certain about which is weakest, see if you can dismiss that option quickly. • If the case has a lot of quantitative evidence, which criterion is most relevant to it?

If you don’t have a criterion relevant to the quantitative evidence, formulate one. Work up as much relevant, high-value quantitative evidence as you can. Do No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC op yo • Think about the frameworks that seem most appropriate to the situation. Quickly review the speci? cs of the frameworks if you aren’t sure of them. rP os t HOW TO ANALYZE A CASE 13 14 ANALYSIS • If there are con? icts about the decision between individuals or groups, think about why that is.

Look at the decision from the point of view of each of the parties to the con? ict. • If the protagonist is in a dif? cult position in relation to the decision, consider why that is. EVALUATION • Review the criteria you have come up with so far. Which do you have the most con? dence in? • Do you already have a sense of the bottom-line evaluation you favor? If you do, what are the reasons for the preference? Pursue those reasons. • Start by applying the criterion that seems to identify the most evidence in the case. • Investigate the most positive rating or the most negative with the criterion you have the most con? ence in. • If the case has a lot of quantitative evidence, which criterion is most relevant to it? If you don’t have a criterion relevant to the quantitative evidence, formulate one. Work up as much relevant, high-value quantitative evidence as you can. Do No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC Taking notes helps you organize and remember information, but it serves the equally important purpose of recording your thought process.

Without note taking, you can too easily stray from active reading. Of course, note taking can degenerate into transferring information in the case to a piece of paper or computer screen. Notes on a case don’t simply record facts. They capture anything that might lead to answers to the questions you’ve asked. It may sound trivial, but I recommend that students try to contain the “ highlighter habit. ” This study aid is well adapted to the lecture model of learning, but it can be a detriment to case study. Highlighting sentences is satisfying because it makes you feel you’re doing something.

In reality, what you’re doing is marking sentences to think about later, and that’s a setup for passive reading. You should be thinking about statements the ? rst time you encounter them. That said, highlighters can be useful as a tool to differentiate related content: facts about one aspect of the case, for example, or text and numbers that belong to one category of evidence. op yo • What are the terms of the evaluation going to be (e. g. , strengths/ weaknesses)? Do any stand out in the case (e. g. , an obvious strength of an individual)? rP os t A pencil or pen is more conducive to active reading—to write down questions and make notes.

When you begin to gravitate toward a conclusion, stop work and write it down. The function of a hypothesis is to give you a position to try out, not a ? nal conclusion, so listen carefully to your intuition. If you have time, put the case away after this iteration. Even a short break can be useful. There is scienti? c evidence that our subconscious minds are much better at dealing with complexity than our conscious minds. Turning your attention to something else allows that subconscious capacity to work on the information you have collected. 4. Proof and Action (40 minutes) Do

A hypothesis drives a different approach to the case. You want to prove something, not look for something to prove. Ask these questions: What evidence do I have that supports the hypothesis? What additional evidence do I need? Look at the information you’ve compiled and identify evidence supporting the hypothesis. Your ? rst priority should be to add to the evidence you have. What is the strongest evidence? Can you add more to it? Now assess where evidence is missing. Where will you ? nd more—or is there any evidence in the case? Think about any factors you may have overlooked such as a cause, criterion, or evaluative category.

Go back into the case, with the single purpose of bringing out more evidence that aligns with your hypothesis. You don’t have to work from the ? rst page to the last. You can go directly to the sections and exhibits you think have what you need. Of course, you can work from beginning to end if that makes you more comfortable. Just be sure to stay focused on what you’re trying to prove. Let’s say that you’re building an argument for a decision option and one of the criteria is cost savings. You’ve noted some statements that imply your decision option will savemoneyfor the ? m and circled numbers that you thought were relevant to savings. Collect those numbers now, and work out calculations to estimate the total savings. You may then have one of those gratifying moments of case study: from those scattered numbers that looked so inconsequential when viewed individually, you’ve pulled together an estimate that indicates a very large annual savings—and that’s just one part of your argument. Also give some thought to the actionable content of your position. How would you implement the decision you’re recommending? What actions does your diagnosis or evaluation call for? Think in practical, eal-world, not ideal-world, terms. Don’t just sketch out in your mind a broad No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC op yo rP os t HOW TO ANALYZE A CASE 15 16 ANALYSIS approach to action. Think about tangible actions and write them down. Finally, give a bit of thought to the order of the actions. An action plan is a program in which actions are taken at a certain time for a reason. It isn’t a to-do list. 5. Alternatives (15 minutes) PROBLEM Can the problem be de? ned differently?

Would that make a difference to the diagnosis? Are there any holes in the diagnosis—could there be causes missing? What’s the weakest part of the diagnosis? Could an entirely different diagnosis be made? What would it look like? DECISION No Do This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC What’s the biggest downside of the recommended decision? How would you manage the downside? What’s the strongest evidence against the recommendation? How would a case for the major alternative look?

EVALUATION Have you been objective and thorough about the evaluative ? ndings that oppose your overall assessment? Think how a different overall evaluation might be proved. Have you accounted for factors that the subject of the evaluation couldn’t control? “ BUT WHAT IF MY HYPOTHESIS IS WRONG? ” Students have asked me that question many times. A hypothesis isn’t wrong; a hypothesis fails when you can’t make a credible argument for it from case evidence. If you ? nd yourself in that situation—and you will sooner or later—? rst make sure the dif? culty lies with the hypothesis and not with your evidence gathering.

You may have overlooked important information or not used specialized tools effectively. If you’re certain the evidence isn’t there, face up to it but realize that the work you’ve already done isn’t wasted. op yo It may seem paradoxical, but the last phase of analysis should be to question your own hypothesis: What is the greatest weakness of the hypothesis? What is the strongest alternative to it? The intention isn’t to undermine yourhard workbut to take a step back and look critically at the hypothesis and the evidence. Every position has a weakness, and you should be the one who recognizes it, not the professor or your peers.

Here are some ways to think critically about your work: rP os t You now have a good grasp of the 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 looking at the fear of being wrong is to ask yourself what the alternative is. I have not heard of a method of case analysis that never leads to dubious conclusions. In fact, making analytic mistakes is invaluable. Through mistakes, we learn more about the thought process called case analysis.

And a shaky analysis can sometimes be a symptom of risk taking, which is also an invaluablelearning experience. 1. Michael E. Porter, Competitive Strategy: Techniques for Analyzing Industries and Competitors (New York: The Free Press, 1980). 2. The American Heritage College Dictionary, third edition (Boston: Houghton Mif? in Company, 1993). Do No This document is authorized for use only by Imtiaz Ahmed until October 2010. Copying or posting is an infringement of copyright.[email protected]harvard. edu or 617. 783. 7860. tC op yo NOTES rP os t HOW TO ANALYZE A CASE 17