# Case Reading Guide: [Course and case title] > A guide to reading and preparing a business case in 90 to 120 minutes, and to taking part in a case discussion: how > the case method works and where it came from, a five-step reading protocol (skim, find the protagonist and the > decision, read the exhibits first, build the analysis, take a position), cold calls, preparation with a learning > team, a practice plan and an original practice mini-case with four exhibits and a teaching note. It is for admitted > students before the programme starts, first-year students in their first weeks of cases, and anyone preparing for > a case competition. Source basis: the public history of the case method (Harvard Business School Baker Library, > Garvin 2003), the student literature on case analysis named under further reading, and the learning-science > evidence on retrieval and spacing (Dunlosky et al. 2013), all written up in original words. Replace every > [bracketed] value and every row marked *Example*. ## Document control | Field | Value | |---|---| | Learner | [Your name] | | Programme and course | [Programme], [course code and name] | | Term | [Term and year] | | Learning team | [Team name or number, members] | | Course policy on AI tools and prior-year materials | [Summary and link to the syllabus or honour code] | | Reviewed by | [Learning-team member, mentor or coach] | | Version | [0.1] | | Last updated | YYYY-MM-DD | ## Contents - [1. How the case method works](#1-how-the-case-method-works) - [2. Types of cases](#2-types-of-cases) - [3. The reading protocol: 90 to 120 minutes](#3-the-reading-protocol-90-to-120-minutes) - [4. Margin codes and the reading log](#4-margin-codes-and-the-reading-log) - [5. Building the analysis](#5-building-the-analysis) - [6. Taking a position](#6-taking-a-position) - [7. In the classroom: discussion and cold calls](#7-in-the-classroom-discussion-and-cold-calls) - [8. Preparing with a learning team](#8-preparing-with-a-learning-team) - [9. Practice mini-case: Northfield Outdoor Co. gear rental](#9-practice-mini-case-northfield-outdoor-co-gear-rental) - [10. Practice plan: 12 sessions](#10-practice-plan-12-sessions) - [11. How case work is graded](#11-how-case-work-is-graded) - [12. Common struggles and mistakes](#12-common-struggles-and-mistakes) - [13. Self-test questions](#13-self-test-questions) - [14. Teaching note for the practice case](#14-teaching-note-for-the-practice-case) - [15. My notes and next steps](#15-my-notes-and-next-steps) - [Sources and further reading](#sources-and-further-reading) - [Answer key](#answer-key) - [Revision history](#revision-history) ## 1. How the case method works **What a case is.** A business case is a written account of a real or realistic situation, usually ending at the moment a named manager (the protagonist) must decide something. It gives you the facts the manager had, often including irrelevant ones and leaving out others you would like, plus exhibits with financial and operating data. There is no answer at the back of the book. Class time is a structured discussion in which the instructor asks students to diagnose the problem, argue for options, test each other's analysis and commit to a decision. **Where it came from.** The method came to business education from law: Harvard Law School began teaching from cases in the 1870s. Harvard Business School, founded in 1908, moved to case-based teaching in the early 1920s under Dean Wallace B. Donham (dean from 1919). Melvin T. Copeland's *Marketing Problems* (1920) was among the early business casebooks, and in 1922 the faculty voted to rename what it had called the "problem method" the "case system" (Harvard Business School Baker Library exhibit on the case method, checked October 2026; histories differ by a year or so on some dates). Garvin (2003) compares how law, business and medical schools use cases. Today the method is used, in different proportions alongside lectures, simulations, projects and experiential courses, across most leading MBA programmes; each school publishes how much of its curriculum is case-based, so check the programme's own pages. **Why it works.** A case asks you to retrieve what you know, apply it to a messy situation, defend a choice in public and revise it when a classmate finds a flaw. Retrieval practice and spaced practice are among the learning techniques with the strongest evidence (Dunlosky et al. 2013), and a term of cases is several hundred rounds of both. The method also builds the competencies business-school accreditation frameworks emphasise: | Competency | How a case builds it | Where you practise it in this guide | |---|---|---| | Analytical and quantitative reasoning | Turning exhibits into a few decisive numbers | Sections 3 (step 3) and 5 | | Ethical decision making | Weighing stakeholders when the profitable option harms someone | Section 5, criteria | | Communication | A two-minute opening, a written recommendation | Sections 6 and 7 | | Leadership and teamwork | Learning-team preparation, building on others in class | Sections 7 and 8 | | Global perspective | Cases set in many countries and industries | Section 10, case selection | | Technology, data and AI | Spreadsheet models, judging data quality, policy on AI tools | Sections 5 and 8 | **What the instructor expects.** You are expected to have read the case closely, done the numbers that matter, and arrived with a position you can defend. You are not expected to be right; you are expected to be prepared, clear and open to changing your mind for a good reason. ## 2. Types of cases Most cases ask for one of four kinds of work. Some combine them; the assignment questions usually tell you which. | Type | The question behind it | What your preparation must produce | *Example* | |---|---|---|---| | Decision | What should the protagonist do? | A choice among options, with criteria and an implementation plan | Roll out a new service to all stores or only some | | Diagnosis | Why did this happen? | A causal explanation supported by evidence from the case | Why did a product launch miss its targets | | Evaluation | How good is this (plan, leader, deal)? | A judgement against explicit standards | Assess a chief executive's turnaround plan | | Illustration or concept | How does this idea work in practice? | The concept stated clearly and mapped onto the facts | A case that shows how a platform's network effects formed | Decision cases dominate most core courses, so the protocol in section 3 is written for them. For a diagnosis case, replace "take a position" with "commit to the main cause"; for an evaluation case, state your standard before you judge. ## 3. The reading protocol: 90 to 120 minutes Read in passes, not front to back. Each pass has a purpose and a time budget for a typical 10 to 20 page case. ```mermaid flowchart TD A["Step 1: Skim (10 to 15 min)"] --> B["Step 2: Protagonist, decision, deadline (5 min)"] B --> C["Step 3: Exhibits first (20 to 25 min)"] C --> D["Step 4: Close read with margin codes (25 to 30 min)"] D --> E["Step 5: Build the analysis (25 to 35 min)"] E --> F["Step 6: Take a position and prepare the opening (10 to 15 min)"] F --> G{"Does the analysis support the position?"} G -->|"Yes"| H["Fill the prep sheet and stop"] G -->|"No"| E ``` | Step | Time | What you do | Output | |---|---|---|---| | 1. Skim | 10 to 15 min | Read the first two and the last two paragraphs, the headings, the exhibit titles and the assignment questions. Do not take notes yet. | A one-line guess at what the case is about | | 2. Protagonist and decision | 5 min | Write who must decide, what, by when, with what authority and constraints | The decision statement in one sentence | | 3. Exhibits first | 20 to 25 min | Go through every exhibit: what it measures, its units and period, what changed, what is surprising. Compute the two or three ratios that matter. | A list of three to five facts the numbers establish | | 4. Close read | 25 to 30 min | Read the text in full with margin codes (section 4). Look for options the protagonist names, constraints, stakeholders and claims that the exhibits contradict. | Coded text, list of options and criteria | | 5. Build the analysis | 25 to 35 min | Answer the assignment questions with numbers. Compare options against criteria. Run one sensitivity on the assumption that matters most. | Analysis table, key numbers | | 6. Take a position | 10 to 15 min | Choose. Write a two-minute opening, the strongest counter-argument and your answer to it, and one question you would ask in class. | The completed mba/mba-case-prep-sheet | **Why exhibits before text.** The text tells the protagonist's story, with the protagonist's biases. The exhibits are closer to evidence. Reading them first lets you notice when the narrative and the numbers disagree, which is often the point of the case. It also means you read the text knowing which facts matter. **When you are short of time.** Do steps 1, 2 and 3, read the last third of the text closely, and take a position anyway. A clear, partly supported position contributes more to a discussion than a complete summary with no view. ## 4. Margin codes and the reading log Use a small, fixed set of margin codes during step 4. On paper use letters; in a PDF reader use highlight colours. | Code | Meaning | What to do with it later | |---|---|---| | D | Decision, deadline or authority | Goes into the decision statement | | O | An option, stated or implied | Goes into the options list | | C | Constraint (money, time, law, capacity, people) | Becomes a criterion or a screen | | S | Stakeholder and what they want | Feeds criteria and implementation | | N | A number in the text not shown in an exhibit | Add it to your own exhibit | | ? | A claim you doubt or that the exhibits contradict | Test it in step 5; good material for class | | ! | A fact that changes your view | Mention it in your opening | **Reading log.** One row per case. After ten cases, look at the pattern: where do you overrun, which codes do you never use, which case types slow you down. | # | Case | Type | Minutes (steps 1 to 6) | Position taken | Called on? | What I missed in class | |---|---|---|---|---|---|---| | 1 | Northfield gear rental *(Example)* | Decision | 12, 4, 24, 28, 34, 12 = 114 | Roll out to the top 18 stores | Yes, cold call | Did not test the "would have bought anyway" figure | | 2 | [Case title] | [Type] | [Times] | [Position] | [Yes/No] | [Lesson] | ## 5. Building the analysis **Start from the decision, not the framework.** Ask what you would need to know to choose between the options, then compute that. A framework earns its place only if it produces a finding (for example, a five forces view that shows supplier power will squeeze the margin the plan relies on, see mba/mba-five-forces-card). **The minimum analysis for a decision case.** 1. **Options.** List every option the case names, plus "do nothing" and at least one option the protagonist has not considered (a phased, partial or staged version is often the best). Three to five options is typical. 2. **Criteria.** Three to five criteria that matter to the decision maker: value created (NPV, profit, payback), risk, strategic fit, feasibility, and effect on stakeholders. Write them before you score, so you do not pick criteria that flatter your favourite. 3. **Key numbers.** The two or three calculations that move the decision. In finance and operations cases this is often an NPV or a break-even (mba/mba-npv-irr-card, mba/mba-cvp-breakeven-card); in marketing a customer or segment economics figure (mba/mba-customer-lifetime-value-card); in strategy a margin or share comparison. 4. **The swing assumption.** The one assumption that would flip your answer if it were wrong. Compute the value at which it flips (the break-even). This single number is often the most valuable contribution you can make in class. 5. **Evidence quality.** Note where data come from: a pilot chosen by enthusiasts, a survey of current customers, a manager's estimate. Biased data are a finding. **Spreadsheet discipline.** Put assumptions in one block with their source ("Exhibit 2", "my estimate"), keep calculations separate, and label units. You will reuse the model in class when someone asks "what if". The Excel drills in mba/pre-mba-excel-modeling-drills train this. ## 6. Taking a position Write your position as a recommendation a busy executive could act on: **Recommendation template**: [Protagonist] should [action, scale and timing] because [reason 1 with number], [reason 2] and [reason 3]. The main risk is [risk]; [mitigation]. I would change my mind if [evidence]. *Example*: Northfield should roll out gear rental to the 18 stores in the top two traffic quartiles next season, not to all 36 remaining stores, because those stores clear the break-even of about 30 rentals a week and add an NPV of about USD 0.49m, while the lower-traffic stores destroy value. The main risk is that the purchase "halo" from renters is smaller than the pilot suggests; a holdout test in the first season will measure it. I would change my mind if the halo turned out to be less than half the pilot estimate. **The opening.** Instructors often open with "what should the protagonist do?" and ask one student to start. Prepare a two-minute answer: decision, recommendation, the two strongest reasons with numbers, the main risk. Do not retell the case: everyone has read it. **Counter-argument.** Write the best case against your position in one sentence and your answer to it. If you cannot answer it, your position may be wrong; that is worth knowing before class. ## 7. In the classroom: discussion and cold calls **How a discussion usually runs.** An opening (one student lays out the situation or a recommendation), then diagnosis, options and analysis, a debate between positions (sometimes a vote), and a wrap-up in which the instructor draws out the general lessons. Instructors use the boards to structure the discussion; watch which board a comment goes on to follow the plan. **Cold calls.** A cold call is when the instructor asks a student who has not raised a hand, often to open the case. Some instructors warn students in advance ("warm calls"); many do not. Expect one, and treat it as routine. | Situation | What to do | |---|---| | Cold-called to open | Give your prepared two-minute opening: decision, recommendation, two reasons with numbers, main risk | | Asked a question you did not prepare | Restate the question, say what you do know from the case, reason aloud to a provisional answer | | Asked for a number you do not have | Say how you would compute it and give an estimate with its basis | | You disagree with the previous speaker | Name their point, state where you differ and why, with evidence from the case | | The discussion has moved on from your point | Hold it unless it changes the direction; otherwise link it explicitly ("going back to the cash point...") | **Types of good contribution.** Opening the case; adding a number that settles a dispute; spotting a contradiction between text and exhibits; building on a classmate's point and taking it further; changing your mind out loud when the evidence warrants it; linking the case to a concept from another course. Repeating a fact from the case, or summarising what has already been said, adds little. **After class.** Write three lines in your participation log (mba/mba-class-participation-log): what you said, what the class concluded, what you would do differently. Note the general lesson the instructor drew; those lessons are what exams test. ## 8. Preparing with a learning team Many programmes assign first-year students to a learning team (also called a study group) of five to seven people with different backgrounds. Teams typically meet before class for 30 to 60 minutes to discuss the day's cases. Set the rules in a team charter (mba/mba-learning-team-charter) in the first week. | Element | Good practice | |---|---| | Individual preparation | Everyone reads the whole case and takes a position before the meeting | | Agenda | 10 minutes per case: each person states a position in one sentence, then the team debates the swing assumption | | Expertise | The accountant checks the numbers, the engineer the operations logic, the marketer the customer view; rotate who leads | | Disagreement | Seek it out; a team that agrees on everything has prepared nobody for class | | Timekeeping | Fixed start and end, a rotating timekeeper | | Feedback | A short monthly round of feedback (mba/mba-peer-feedback) | **Rules on materials and tools.** Many programmes forbid using notes, analyses or teaching notes from earlier years of the same course, and treat it as an honour-code matter. Courses also differ on AI tools: some allow them for background reading or spreadsheet help, some forbid them for case preparation or graded write-ups. Read the syllabus and honour code, record the rules in the Document control table, and ask the instructor when in doubt. Whatever the rules, your class contribution must come from your own analysis: an argument you did not build is one you cannot defend under questioning. ## 9. Practice mini-case: Northfield Outdoor Co. gear rental **Situation.** On a Monday in late September, Adaeze Mensah, vice president of stores at Northfield Outdoor Co., had three weeks to bring a recommendation to the executive committee. Northfield, a mid-size outdoor-gear retailer and brand, ran 40 stores and an e-commerce site. Its stores competed with online sellers on price and with specialist shops on advice, and store traffic had been flat for two years. In March, Mensah had launched a gear-rental pilot in four stores: tents, backpacks, snowshoes and climbing packs rented by the weekend. The regional managers chose the stores; all four sat near popular trailheads and in the top quarter of the chain by foot traffic. The pilot ran for 26 weeks. Customers liked it: satisfaction scores for renters were the highest of any service Northfield had offered, and staff reported that renters often came back to buy the gear they had tried. Mensah's head of finance, Tomas Lindqvist, had built the store-level economics (Exhibits 1 and 2). Each store needed a rental fleet costing USD 38,000, which would last about three years, plus a part-time rental associate for 30 hours a week. Rentals ran for about 40 weeks a year; the deep winter and the wettest weeks were quiet. Lindqvist used Northfield's 10% hurdle rate for store investments and, to keep the first pass simple, ignored taxes and any resale value of the used fleet. The marketing team added data from the loyalty programme (Exhibit 3): 18% of renters bought something from Northfield within 90 days of their rental, spending USD 210 on average at Northfield's 45% gross margin. A follow-up survey asked those buyers whether they would have made the purchase without renting first; 40% said yes. The debate in the executive committee had already started. The chief marketing officer, Lucía Ferreira, wanted to roll out rental to all 36 remaining stores before the spring season: "This is the first thing in years that brings people into the stores. If we wait a year, someone else will own rental." The chief financial officer, Kwame Asante-Boateng, was sceptical: "Four hand-picked stores near trailheads tell us nothing about a suburban mall store. And I am not paying for a fleet on the strength of a survey." He proposed extending the pilot to eight more stores, chosen to represent the chain, for one more season, which would cost USD 304,000 in fleets. Mensah asked her analyst for a view of the rest of the chain. The analyst grouped the 36 remaining stores into four traffic quartiles of nine stores each and, using foot-traffic data and the pilot's ratio of rentals to visits, estimated the rentals each group could expect relative to the pilot average (Exhibit 4). One store in the pilot, Cedar Point, had disappointed: it sat in a smaller town and its rentals had stayed low all season. Mensah thought about the three paths in front of her: a full rollout to all 36 stores, an extended pilot, or something in between. She also wondered how much of the case for rental rested on the purchases renters made afterwards, and how sure Northfield could be of that number. The committee would expect a clear recommendation, a number it could hold her to, and a plan for what would happen if the numbers disappointed. **Exhibit 1. Pilot results, March to August (26 weeks)** *(Example)* | Store | Traffic quartile | Rentals per week (average) | Average fee per rental (USD) | Direct cost per rental: cleaning, repair, loss (USD) | Associate hours per week | |---|---|---|---|---|---| | Ridgeport | Top | 62 | 34 | 9 | 30 | | Lakeview | Top | 48 | 34 | 9 | 30 | | Harlan Falls | Top | 35 | 34 | 9 | 30 | | Cedar Point | Top | 21 | 34 | 9 | 30 | | Average | | 41.5 | 34 | 9 | 30 | **Exhibit 2. Store-level investment and cost assumptions (finance team)** *(Example)* | Item | Value | |---|---| | Rental fleet per store | USD 38,000, three-year life, no resale value assumed | | Rental associate | 30 hours a week at USD 22 an hour, during active weeks | | Active rental weeks per year | 40 | | Hurdle rate for store investments | 10% a year | | Taxes | Ignored in this first pass | **Exhibit 3. Renter purchase behaviour (loyalty data and survey)** *(Example)* | Measure | Value | |---|---| | Renters who bought from Northfield within 90 days | 18% | | Average purchase by those renters | USD 210 | | Gross margin on that merchandise | 45% | | Buyers who said they would have bought anyway | 40% | **Exhibit 4. Remaining 36 stores by traffic quartile (analyst estimate)** *(Example)* | Quartile | Stores | Expected rentals as % of pilot average | Expected rentals per week | |---|---|---|---| | Top | 9 | 100% | 41.5 | | Second | 9 | 80% | 33.2 | | Third | 9 | 60% | 24.9 | | Bottom | 9 | 40% | 16.6 | | All 36 stores (simple average) | 36 | 70% | 29.05 | **Discussion questions.** 1. Who is the protagonist, what exactly must she decide, and by when? What are her options, including any not named in the case? 2. What does one rental contribute, before and after the "halo" from later purchases? Which figure in Exhibit 3 do you trust least, and why? 3. What is the NPV of rental in each pilot store and in a store at the chain average? How many rentals a week does a store need for the fleet to earn the hurdle rate? 4. What does Exhibit 4 change about the full-rollout proposal? Is the pilot evidence biased, and in which direction? 5. What do you recommend to the executive committee, and what result in the first season would make you stop or expand? ## 10. Practice plan: 12 sessions | Session | Focus | Material | Exercise | Deliverable | |---|---|---|---|---| | 1 | How the method works | Sections 1 and 2 of this guide | Classify three case abstracts by type | Three one-line case-type labels | | 2 | Skim and decision statement | Any decision case | Steps 1 and 2 only, 20 minutes | Decision statement | | 3 | Exhibits first | The Northfield practice case | Step 3 with a timer | Three facts the exhibits establish | | 4 | Full protocol | The Northfield practice case | Steps 1 to 6 in 120 minutes | Completed mba/mba-case-prep-sheet | | 5 | Debrief | Section 14 teaching note | Compare your analysis with the note | Error-log entries (mba/mba-error-log) | | 6 | Numbers under time | Accounting or finance case | Key-number analysis only, 40 minutes | NPV or break-even with Excel formulas | | 7 | Diagnosis case | Operations or organisational case | Commit to a main cause | Causal chain with evidence | | 8 | Evaluation case | Leadership or strategy case | State the standard before judging | Evaluation table | | 9 | Speaking | Any prepared case | Record a two-minute opening, listen back | Opening script, revised | | 10 | Team preparation | Case prepared by a study partner too | 30-minute team meeting with agenda | Team disagreements list | | 11 | Speed | Two cases in one evening | Protocol at 75 minutes each | Reading-log times | | 12 | Review | All logs | Find your three most common misses | Personal case checklist | ## 11. How case work is graded In case-based courses, class participation often carries a substantial share of the grade, alongside written case analyses, group projects and exams; the weights differ by school and course, so check the syllabus. Instructors usually judge the quality of contributions over the term rather than their number. A typical participation rubric looks like this *(Example)*: | Level | Description | |---|---| | Excellent | Contributions move the discussion: decisive analysis, new evidence, well-reasoned disagreement, links to concepts; listens and builds on others | | Good | Regular, relevant contributions supported by case facts; some analysis; responds well when called on | | Adequate | Prepared when called on, but contributions mostly restate facts or repeat earlier points | | Weak | Unprepared when called on, absent without notice, or contributions that do not engage with the case | Written case analyses (see mba/mba-case-write-up and academic/finance-case-analysis) are usually graded on the clarity of the recommendation, the quality and correctness of the analysis, the use of case evidence, the treatment of alternatives and risks, and the writing. Case exams give you a new case and a fixed time, so the speed you build with the protocol counts directly. ## 12. Common struggles and mistakes 1. **Reading front to back, once.** You reach the exhibits tired and with no time to compute. Use passes. 2. **Summarising instead of deciding.** The case is not a report to summarise; the deliverable is a decision. 3. **Losing the protagonist.** Analysing "the company" in general instead of what this person can do by this date. 4. **Too many frameworks.** A grid filled for its own sake; use a framework only when it produces a finding. 5. **Taking the case's data at face value.** Pilots chosen by enthusiasts, surveys of happy customers and managers' forecasts are evidence with a bias; name it. 6. **No "do nothing" option and no hybrid.** The best answer is often a phased or partial option nobody in the case proposed. 7. **No swing assumption.** Arriving without knowing which assumption would change your answer, so you cannot respond when a classmate challenges it. 8. **Over-preparing one case and skipping the next.** Consistent preparation across all cases beats perfection on one. 9. **Waiting for the perfect comment.** Early in the term, open with a modest, prepared contribution; confidence follows. 10. **Talking at length.** Long contributions crowd out others and are hard to follow; make one point, with evidence. 11. **Ignoring stakeholders and ethics.** A profitable option that harms employees, customers or the community needs that cost stated and weighed, not left out. 12. **Not closing the loop.** Not writing down the lesson after class, so the same mistakes recur. ## 13. Self-test questions 1. What is the difference between a decision case and a diagnosis case, and how does it change your preparation? 2. Why read the exhibits before the full text? 3. Write the five elements of a decision statement. 4. What is a "swing assumption", and why is its break-even value a strong class contribution? 5. Name three ways a case's evidence can be biased. 6. You are cold-called on a question you did not prepare. What do you do? 7. What should a learning team do before its meeting, and what should the meeting focus on? 8. Which two sources should you check before using AI tools or earlier-year notes for case preparation? 9. In the practice case, what does one rental contribute including the halo, and what is the break-even number of rentals per week? 10. In the practice case, why is the chain-average store a misleading basis for the rollout decision, and what is the NPV of rolling out to the top two quartiles only? ## 14. Teaching note for the practice case **Teaching objectives.** (1) Practise the protocol on a decision case with consistent exhibits. (2) Show that a pilot chosen from the best stores overstates the average result. (3) Show that averages hide the distribution: a rollout that destroys value on average can create value when targeted. (4) Identify the swing assumption (the purchase halo) and design a test for it. **Decision statement.** Adaeze Mensah must recommend within three weeks whether, where and how fast to roll out gear rental to Northfield's 36 remaining stores, within a 10% hurdle rate for store investments. **Unit economics.** Contribution per rental is USD 34 - 9 = 25. The halo is $0.18 \times 210 \times 0.45 \times (1 - 0.40) = 10.21$ per rental (gross margin on purchases that would not have happened otherwise). Weekly store cash flow is rentals $\times$ (25 + 10.21) less USD 660 for the associate; over 40 weeks the associate costs USD 26,400 a year. **Store NPVs.** Annual cash flow over three years, discounted at 10%, less the USD 38,000 fleet. Excel: `=NPV(10%,CF,CF,CF)-38000`, where CF is the annual cash flow. | Store or group | Rentals per week | Annual cash flow (USD) | NPV (USD) | Payback (years) | |---|---|---|---|---| | Ridgeport | 62 | 60,911 | 113,476 | 0.62 | | Lakeview | 48 | 41,196 | 64,447 | 0.92 | | Harlan Falls | 35 | 22,888 | 18,920 | 1.66 | | Cedar Point | 21 | 3,173 | -30,109 | 11.98 | | Pilot average | 41.5 | 32,042 | 41,684 | 1.19 | | Chain average (70%) | 29.05 | 14,509 | -1,917 | 2.62 | **Break-even.** The equivalent annual cost of the fleet is USD 15,280 (`=PMT(10%,3,-38000)`). The rentals a week needed for NPV = 0 are $(15{,}280 + 26{,}400) / (35.21 \times 40) = 29.6$ with the halo, and 41.7 without it. The pilot average (41.5) does not clear the hurdle on rental income alone: the case for rental rests on the halo. **Rollout options.** | Option | Stores | Fleet investment (USD) | NPV (USD) | Comment | |---|---|---|---|---| | A. Full rollout | 36 | 1,368,000 | -69,024 | Third and bottom quartiles destroy USD 557,724 | | B. Extended pilot | 8 | 304,000 | Not estimated | Buys information, delays the strongest stores by a year | | C. Top two quartiles | 18 | 684,000 | 488,699 | Top quartile 375,153, second 113,547 | By quartile, NPV is USD 375,153 (top, 41.5 rentals a week), 113,547 (second, 33.2), -148,059 (third, 24.9) and -409,665 (bottom, 16.6). The chain average of 29.05 rentals sits just below the break-even of 29.6, so an analysis on the average store says "slightly negative, do not proceed", while the targeted rollout is clearly positive. **Sensitivity to the swing assumption.** | Halo assumed | Halo per rental (USD) | Break-even rentals per week | NPV, top 18 stores (USD) | NPV, all 36 stores (USD) | |---|---|---|---|---| | As measured | 10.21 | 29.6 | 488,699 | -69,024 | | Half | 5.10 | 34.6 | 147,429 | -599,889 | | None | 0.00 | 41.7 | -193,841 | -1,130,754 | The targeted rollout survives a halo half as large as measured; it does not survive no halo at all. The halo figure is also the weakest evidence in the case: the 40% "would have bought anyway" comes from a self-report survey, renters in trailhead stores are likely to be keen buyers anyway, and the 90-day window may pull forward purchases that would have happened later. **One defensible recommendation.** Roll out to the 18 stores in the top two quartiles before spring (Option C), keep a small holdout group of comparable stores or customers to measure the true halo in the first season, and set a rule: expand to the third quartile only if measured incremental purchases support a break-even below its expected 24.9 rentals a week; close rental in any store below about 30 rentals a week after two seasons (Cedar Point is the first candidate). This answers the CMO's concern about timing and the CFO's concern about evidence. **Discussion plan (80 minutes).** Opening and decision statement (10); unit economics and the halo (15); store NPVs and the break-even (15); the pilot bias and Exhibit 4 (15); options vote and debate (15); lessons: pilots, averages, swing assumptions (10). ## 15. My notes and next steps | Date | Case | What went well | What I will change | Template to use next | |---|---|---|---|---| | YYYY-MM-DD | Northfield gear rental *(Example)* | Exhibits pass found the pilot bias | Compute the swing assumption's break-even earlier | mba/mba-case-prep-sheet | | YYYY-MM-DD | [Case] | [Note] | [Change] | [Template] | ## Sources and further reading - Baker Library, Harvard Business School. *Business Education and the Case Method* (online exhibit), library.hbs.edu, checked October 2026. - Barnes, L. B., Christensen, C. R., and Hansen, A. J. (1994). *Teaching and the Case Method* (3rd ed.). Harvard Business School Press. - Copeland, M. T. (1920). *Marketing Problems*. A. W. Shaw Company. - Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., and Willingham, D. T. (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. *Psychological Science in the Public Interest*, 14(1), 4-58. - Ellet, W. (2018). *The Case Study Handbook: A Student's Guide* (revised edition). Harvard Business Review Press. - Garvin, D. A. (2003). Making the case: Professional education for the world of practice. *Harvard Magazine*, September-October 2003. Open resource: OpenStax, *Principles of Management* (CC BY-NC-SA 4.0), openstax.org, for background on the management concepts cases draw on. ## Answer key 1. A decision case asks what the protagonist should do, so preparation ends in a recommendation with options and criteria. A diagnosis case asks why something happened, so preparation ends in a causal explanation supported by evidence; forcing a recommendation onto it misses the task. 2. The exhibits are closer to evidence than the narrative, reading them first shows you which facts matter, and it lets you spot where the protagonist's story and the numbers disagree. 3. Who decides, what, by when, with what authority, under which constraints. 4. The assumption that would flip your answer if wrong. Its break-even value tells the class exactly how wrong it would have to be, which turns a debate of opinions into a test of evidence. 5. Any three of: a pilot or sample chosen by enthusiasts or from the best sites; self-reported survey answers; a manager's own forecast; survivorship (only successful units reported); a short or unusual time period. 6. Restate the question, say what you know from the case, reason aloud to a provisional answer, and give an estimate with its basis if a number is asked for. Saying how you would find out is better than silence. 7. Each member prepares the whole case individually and takes a position. The meeting focuses on positions and disagreements, especially the swing assumption, not on dividing up the reading. 8. The course syllabus (including its AI policy) and the school's honour code; ask the instructor when unclear. 9. USD 25 rental contribution plus a halo of USD 10.21 gives USD 35.21 per rental. Break-even is 29.6 rentals a week with the halo (41.7 without). 10. The average (29.05 rentals a week) mixes stores far above and far below break-even, and sits just below it, so it hides that the top two quartiles create value while the bottom two destroy it. Rolling out to the top 18 stores has an NPV of USD 488,699 at the measured halo. ## Revision history | Version | Date | Author | Change summary | |---|---|---|---| | 0.1 | YYYY-MM-DD | [Your name] | First copy for [course] | | 0.2 | YYYY-MM-DD | [Your name] | *Example:* Cut step 4 to 20 minutes after ten cases, added code "N" | ## Standard and licence note Structure aligned with the case method as described in the public history of business education (Harvard Business School Baker Library exhibit, Garvin 2003) and the student literature named above, with learning design from Dunlosky et al. (2013). The protocol, the examples, the practice mini-case, its exhibits and its teaching note are original; no published case, teaching note or textbook text is copied or paraphrased, and Northfield Outdoor Co. and every person in the case are fictional. Schools are named only to describe the history of the method; this template is independent and not affiliated with or endorsed by any school or case publisher. Check your own course syllabus and honour code for the rules that apply to you. Template licence: MIT.