Case interview questions, with strong answers
12 questions from Compass's mock interviews, the way an interviewer asks them. Each has a strong answer, a common miss, and why one lands and the other doesn't.
Structure the case
Our client is a regional coffee chain with 200 stores. Profits fell 15% last year while sales were flat. How would you approach this?
A strong answerFlat sales and lower profit point to costs or mix. I'd split profit into revenue and costs, check price and volume by product, then fixed and variable costs.
Structured, and it uses the clue in the prompt: flat sales points you at costs and mix.
A common missI'd start by recommending a price increase, since coffee customers are loyal and a small rise on every cup would win back the profit quickly.
A fix before a diagnosis. Find where the profit went first.
Good. Where would you like to start, and what data would you ask me for?
A strong answerCosts over the last three years, split into fixed and variable, per store. If variable cost per cup rose, I'd then ask what's inside it: beans, milk, labor.
Specific, and it says what you'd do with the answer.
A common missI'd ask for customer satisfaction surveys first, because if people are unhappy with the coffee that would explain why the profits have dropped.
Sales are flat, so customers aren't the first suspect. Follow the clue.
Variable cost per cup went from $1.10 to $1.40, mostly labor. Fixed costs are flat. What does that tell you?
A strong answerLabor per cup is up about 25%. With flat volume, either wages rose or each cup takes longer. I'd ask about wage changes and about the menu mix shifting to slower drinks.
Quantified, and two testable causes. That's the next branch.
A common missIt tells us fixed costs aren't the issue, so the problem must be the rent on the stores, which has probably gone up in most of their locations.
Rent is a fixed cost, and fixed costs were flat. Check the data again.
Wages are flat. Cold drinks with custom toppings went from 20% to 45% of orders and take twice as long to make. Give me your recommendation.
A strong answerSlow drinks are priced like fast ones. Reprice custom cold drinks for their labor, streamline the toppings station, and track labor per cup by drink.
Answer first, tied to the cause, with a way to check it worked.
A common missThe client should stop selling cold drinks with toppings, since they take twice as long, and focus its menu on the hot coffee it's known for.
Cutting 45% of orders throws away demand. Fix the price or the process.
Market sizing
A client wants to launch e-bike rentals in a US city of 1 million people. How big is the yearly market?
A strong answerI'll go top-down: adults, the share who'd rent an e-bike, trips per renter per year, and price per trip. Then I'll check it bottom-up from a fleet.
Approach first, then a cross-check. Exactly how to open a sizing.
A common missI know from the news that bike sharing is a multi-billion dollar industry, so I'd take a small slice of that number for one city as the answer.
Borrowing a headline isn't sizing. Build it from people or bikes.
Walk me through your assumptions.
A strong answerAbout 800,000 adults; 5% rent at least monthly, so 40,000 riders; 20 trips a year each; $4 a trip. That's 40,000 × 20 × $4, about $3.2 million.
Each assumption is stated, rounded and multiplied cleanly.
A common missEveryone in the city could rent, so 1 million people, twice a week each, at $4 a trip. That's 1 million × 104 × $4, so about $416 million a year.
Every resident renting twice a week fails a sanity check.
How do you know $3.2 million is reasonable?
A strong answerBottom-up: 500 bikes, 3 trips a day, $4 a trip, 300 days is about $1.8 million. Same order of magnitude, so call it $2 to $3 million.
A second route that lands close. That's what builds trust in the number.
A common missI'm confident in it because I was careful with every step, so I don't think the number needs any further checking before we use it.
Careful steps still hide bad assumptions. Always cross-check.
So should the client launch?
A strong answer$2 to $3 million a year is small, so it only works with low fleet costs. I'd launch only if a 500-bike fleet pays back within 2 years; next I'd size costs per bike.
Turns the number into a decision and names the next step.
A common missYes, because $3.2 million is a lot of money, and any market worth millions of dollars a year is worth entering for a company of this size.
Revenue isn't profit. A small market can still lose money.
Read the chart, land the answer
A pharmacy chain with 900 stores is considering a telehealth service. Should they launch it?
A strong answerFour questions: is the market attractive, can a pharmacy beat telehealth startups, does it make money, and can its pharmacists run it?
Tailored branches, each a question you can answer with data.
A common missI think they should launch, because telehealth is growing fast and every big healthcare company is moving into it right now, so they can't wait.
A conclusion before the structure. Earn it with the analysis.
Here's survey data: 60% of their customers would try telehealth, but only 15% would pay more than $30 a visit. Startups charge $79. What do you take from it?
A strong answerInterest is high but willingness to pay is far below $79. They'd compete on price, so it works only if a visit costs them under $30.
The so-what, plus the condition that decides it.
A common missThe 60% shows the demand is there, so the chain should match the $79 price of the startups and enter the market as soon as possible.
Only 15% would pay $30. At $79 the demand mostly disappears.
If a visit costs them $25 and they charge $29, how many visits a year do they need to cover $20 million of fixed costs?
A strong answerThe margin is $4 a visit, so $20 million divided by $4 is 5 million visits a year, which is about 5,500 visits per store, around 15 a day.
Clean math, then translated per store to test if it's realistic.
A common missThey make $29 a visit, so $20 million divided by $29 is about 690,000 visits a year, which is very achievable across 900 stores.
Use the margin, not the price. Costs eat $25 of each $29.
Wrap up. What do you tell the CEO?
A strong answerPilot it, don't roll it out. It breaks even at about 15 visits per store a day at $29. Risk: low uptake. Next: 50 stores for 6 months.
Answer first, numbers as reasons, a risk and a concrete next step.
A common missWe looked at the market, then the survey, then the economics. The market is big, the survey was mixed, and the economics depend on volume.
A recap, not a recommendation. Lead with the answer.
Reading answers helps. Saying them under a little pressure helps more: Compass runs these as mock interviews and scores each turn.
Practice freeMore interview guides
- The profitability framework for case interviews22-second explainer with a worked example.
- MECE issue trees for case interviews, with an example23-second explainer with a worked example.
- Market sizing in case interviews: top-down or bottom-up23-second explainer with a worked example.
- Product manager interview questions, with strong answersProduct sense, metrics and strategy questions from real product decisions, each with a strong answer, a common miss, and why one works.
- Technical program manager interview questions, with strong answersTPM interview questions on planning, dependencies, risk and launches, each with a strong answer, a common miss, and why one works.
- Sales and account executive interview questions, with strong answersAccount executive interview questions and role-plays on discovery, pipeline and closing, each with a strong answer, a common miss, and why.
- Investment banking interview questions, with strong answersInvestment banking technical questions on accounting, valuation and deals, each with a strong answer, a common miss, and why one works.