Short answer: case study interview questions come in six families — clarifying, framework, numbers, judgment call, recommendation and pushback. Answer each with four moves: restate and aim, signpost a structure, fill it with evidence out loud, land the point, name its limit. Every example on this page is invented for practice.
Warm up with the timed question drill
Reading about answer methods feels productive and changes almost nothing. The skill that scores is producing a structured sentence in under five seconds while someone watches. So the drill comes first: 12 invented prompts in flashcard form. Pick a question type, run the 90-second timer, answer out loud, then flip the model outline and compare. Mark each card answered and the session keeps count; filter and progress survive a reload; Reset wipes the lot.
Two rules make the practice real. Answer spoken, never in your head — the drill tests what a listener hears. When the timer hits zero, finish the sentence you are on, then stop: the habit being built is landing a point under pressure, not narrating until interrupted.
Timed question drill — 12 cards
Filter by type, run the 90-second clock, answer aloud, compare with the outline. Nothing leaves this page.
Card 1 of 12 in play · all types
"Our regional airline is thinking about charging for seat selection. Your read?"
Illustrative mini-case: a 30-plane regional carrier; two rivals already charge; no data pack given.
- Ask the one clarifying question that changes the answer: is the goal new revenue or calmer boarding?
- Restate what the case already gives so the room hears your base.
- Point forward: "I would start from the frequent-flyer share, because that group decides whether the fee backfires."
"The council wants our museum open until 9 pm. Should it be?"
Illustrative mini-case: a free-entry city museum; the gift shop covers 6% of costs; two funded late nights a week.
- One question first: is success counted in visits, revenue, or access?
- State the assumption you will run with meanwhile — visits first, since the money is public.
- Name the evidence that settles it: hour-by-hour footfall from a one-month trial. Keep it near 20 seconds.
"A 14-screen cinema's profits fell two quarters in a row. Find the leak."
Illustrative mini-case: admissions flat; average concession spend per head down; rent unchanged.
- Split it in one breath: revenue is admissions times spend per head; costs sit underneath.
- Follow the case — admissions flat, so the leak is per-head spend. Say so out loud.
- Mark one driver per branch (snack prices, staffed-lane queues, app discounts) and the till data that settles it.
"A ferry operator is weighing a winter route. Structure the decision."
Illustrative mini-case: a seasonal island route; the vessel sits idle all winter; demand unknown.
- Build the split: winter demand, running cost versus idle cost, and the fare the route could carry.
- Use the idle-ship fact: the honest comparison is marginal cost against winter revenue, not full-year accounting.
- Flag the dominating unknown (demand), then end on a test — a three-month trial under a cost cap.
"A regional zoo is losing school-group bookings. What is your structure?"
Illustrative mini-case: school groups used to be 40% of weekday visits; bookings down a third across two years.
- Split demand by visitor type — schools, families, tourists — and let the case point at the school branch.
- Work that branch: curriculum fit, transport cost, timing against the school year.
- Aim at a decision — what the zoo changes — not a description of the decline.
"Estimate daily demand for bicycle parking at a new rail station."
Illustrative mini-case: a suburban station projected at 4,000 weekday boardings; no cycling data.
- Announce the build before the arithmetic: boardings, times the share who cycle, times the share who park.
- Give every assumption a reason out loud — commuters park all day, off-peak riders mostly do not.
- Keep numbers round, show each step, and sanity-check against a comparable station.
"A $9,000 binding machine cuts unit cost from $1.10 to $0.40 on 5,000 booklets a year. Worth it?"
Illustrative mini-case: the shop outsources binding today; machine upkeep runs $400 a year.
- Write the per-unit saving first: $0.70 a booklet — say it before multiplying.
- Run the year: 5,000 × $0.70 = $3,500, minus $400 upkeep — under half the machine cost in year one.
- Name what would flip it, then answer: not this volume; revisit above about 13,000 booklets.
"Average basket fell from $38 to $34 while transactions rose 8%. What happened to revenue?"
Illustrative mini-case: a grocery chain's quarterly figures, nothing else supplied.
- Set the relationship first: revenue equals transactions times basket.
- Index the old quarter: 100 × 38 = 3,800; the new one is 108 × 34 = 3,672.
- Give the direction before the decimals — down about 3.4% — then name what you would check next.
"The data is thin and the board meets tomorrow. Pick the program to cut."
Illustrative mini-case: three after-school programs, similar costs; one strong on surveys but weak on attendance, another the reverse.
- State the decision rule before the pick: cut the program whose claimed outcome has the weakest evidence.
- Make one pick, plainly — no hedging chorus.
- Name the risk you accept, the cheap check that would catch it, and the two-week upgrade if bought time.
"Your numbers contradict the most senior person in the room. Do you speak?"
Illustrative mini-case: a seminar role-play — your spend-per-head figure disagrees with the director's memory of last year.
- Yes — and aim at the number, never the person.
- Show the figure and its source, then ask for theirs: which year, which definition?
- Offer the reconciliation — two definitions, two years — and remember composure is what is scored.
"One move for the tutoring company: expand online, or open a second city?"
Illustrative mini-case: 300 students in one city; a waitlist of 40; tutors stable.
- Land the pick in the first sentence — no suspense arc.
- Tie two reasons to case facts: the waitlist proves local demand; online adds reach without a lease.
- Reject the other option for a named reason, then end with a step and a number: pilot the 40 waitlist students this term.
"The bank wants a yes or no on an app-only savings product. Land it."
Illustrative mini-case: a regional bank with 90,000 depositors, median age 51; 22% use the app monthly.
- Give the verdict up front: a qualified yes — pilot-sized, not a full launch.
- Anchor it in the case fact: the median depositor is not the app cohort, and a full launch bets the brand on a minority.
- Name the risk (quietly moving branch deposits), the guardrail, and a decision the room can vote on today.
Answered 0 of 12 cards.
Weak on one family? Filter to it and run the cards twice — blocked practice beats shuffling. And if the questions you fear most are about a deck you built yourself, that is a different sport: the Q&A scripts in how to present a case study cover defending your own slides.
The six question types and what each one tests
Nearly every live case question — seminar or hiring round — belongs to one of six families. Learning them pays twice: you spot the type in the first clause, and the type picks the opener. The phrases below are illustrative openers, not scripts.
| Question type | What it tests | Opener phrase | The trap |
|---|---|---|---|
| Clarifying | Whether you aim before you fire | "Before I answer — is the goal X or Y?" | Collecting questions instead of asking the one that changes the answer |
| Framework | Whether you can structure an unfamiliar problem on the spot | "I will split this into demand and cost…" | Naming a famous framework instead of using one |
| Numbers | Whether you reason with figures out loud | "Roughly — I will build it up in pieces…" | Silent mental math, then a number with no visible working |
| Judgment call | Whether you choose when the data will not decide for you | "The case forces a choice, so I would take X…" | Refusing to choose, or choosing without a decision rule |
| Recommendation | Whether you can land a decision, not just analysis | "I recommend X, for two reasons…" | A "recommendation" that just re-lists the analysis |
| Pushback | Whether you defend a position without collapsing or digging in | "Fair challenge — the difference is…" | Defending the conclusion instead of naming the condition it rests on |
Order matters as much as the list. Clarifying comes first even when the question is really a recommendation, because one good question buys five seconds and a narrower target. Pushback comes last and is earned — you cannot be pushed back from an answer you never landed.
A four-move answer that fits any type
Whichever opener you pick, the answer underneath runs the same four moves:
Restate and aim
Say the question back in your own words and name the decision inside it. This is where a clarifying question lives if you need one — and where half your thinking time goes without anyone noticing.
Signpost the structure
Announce the two or three buckets before filling them, so every fact lands in a slot the listener already built. For practice with the buckets themselves, a SWOT analysis generator or a quick PESTLE analysis keeps the common splits sharp.
Fill with case evidence, out loud
One fact per bucket, numbers in round terms, arithmetic in the open. If the case lacks the fact you want, say what you would collect and move — that sentence earns credit everywhere except in your own head.
Land the point and name the limit
One-sentence takeaway, then the honest boundary: what you would check with another day. The limit line separates a strong answer from a lucky one, because it shows the conclusion is conditional, not accidental.
That skeleton mirrors how a written analysis is graded — claim, structure, evidence, limits. The written version is covered step by step in how to write a case study, with a worked model of the analysis layer in case study analysis example.
Eight more example questions with model outlines
The drill above holds 12 prompts; here are 8 more to read, then attempt from memory. Every scenario is invented for this page and labeled illustrative. Cover the answer, say your own out loud, compare — the comparison is where the learning happens.
"Should the online course platform add a paid career-coaching tier?" (Clarifying)
Illustrative: a subscription platform, support tickets full of job-search questions, no coaching staff. Outline: ask which segment is asking — final-year students or career switchers, because pricing differs; restate the revenue model the case gives; name the evidence to gather (ticket themes, willingness-to-pay); hand the floor back inside 20 seconds.
"Delivery times slipped below the 30-minute promise. Diagnose." (Framework)
Illustrative: a pizza chain, promise printed on the box, complaints up in two suburbs. Outline: split the delivery run into kitchen, dispatch and route; ask which stage the case implicates before theorizing; one driver per stage (order batching, driver shifts, road works); finish with the log you would pull — dispatch timestamps by hour and zone.
"A toy maker sources 70% of output from one factory. What is the risk structure?" (Framework)
Illustrative: a single overseas factory, holiday-season peak, no second supplier qualified. Outline: structure as likelihood, impact and mitigation cost; quantify the outage scenario in weeks of peak production; ladder the options from dual-sourcing to buffer stock to contract clauses; recommend the cheapest lever that covers the worst credible case. A Porter's five forces generator covers the supplier-power branch when the question widens to industry structure.
"Estimate annual running-shoe sales in a city that hosts a marathon." (Numbers)
Illustrative: a city of one million, marathon weekend each October. Outline: population, times the share who run at all, times pairs per year, plus the event bump; keep every figure round and spoken; sanity-check against the running shops you can name; land on a range, not a number — ranges show you know the precision you lack.
"6,000 subscribers, 300 added a month, 5% monthly churn. Where does it land in a year?" (Numbers)
Illustrative: a subscription-box business, figures as given. Outline: spot the near-equilibrium first — 5% of 6,000 is 300, so inflow and outflow cancel; say that before computing anything; note churn shrinks as the base shrinks; answer: roughly flat near 6,000 unless a rate moves, and name which one you would move.
"Two after-school programs score the same. The budget funds one. Decide." (Judgment)
Illustrative: equal test-score gains; one is cheaper but newer; the other has three years of data. Outline: state the rule — when outcomes tie, decide on evidence quality and reversibility; pick the program you could unwind cheaply if the three-year advantage was an artifact; name what would flip the pick; resist reopening a decision the room heard you close.
"A rival just cut prices 15%. Recommend the response." (Recommendation)
Illustrative: a mid-size grocery chain facing a discounter's entry. Outline: do not match blindly — say the cost of matching in margin points; segment customers by price sensitivity; recommend the targeted move (match on staples only, or hold price and bundle service); set the guardrail metric and review date; one sentence, one decision, no menu of maybes.
"We tried that two years ago and it failed. Respond." (Pushback)
Illustrative: your late-opening recommendation meets a failed 2024 trial. Outline: agree first — "you did, and that trial matters"; ask what failed specifically, because the answer usually names the condition your version changes; state the changed condition and the small test that checks it; defend the condition in the case, never the conclusion in the abstract.
Cold calls, open discussion and hiring rounds
The same six families arrive in three settings, and each scores a different thing. In a seminar cold call, the professor wants the room to see a method work, so structure and fair-mindedness carry the grade; 30 to 60 focused seconds beat a two-minute tour. In open discussion, the tested move is building on a peer — "yes, and the cost side of that is…" — the same skeleton with a nod to what was just said. In hiring-style rounds, follow-ups dig deeper, so arithmetic stays visible, assumptions get labeled, and composure under interruption counts.
Participation grades make the seminar version worth real points. Find the weight before deciding how much to prepare: the final grade calculator turns "10% for participation" into the actual points at stake. For the reading habits that make cold calls survivable, the three-pass method in how to prepare for business case studies is the fastest fix, and if the discussion ends in a graded talk, the case study presentation rubric scores it the way graders do. The written foundation under all three settings is decoded in how to write a business case study analysis, and if the genre's vocabulary is fuzzy, what is a business case study answers the twelve questions students actually ask.
The eight-point practice checklist
Run this weekly in drill season — 20 minutes that cover more ground than an hour of silent rereading.
- One family at a time — filter the drill to your weakest type and run all its cards.
- Spoken, always — every answer out loud, timer running. Silent practice trains nothing a listener sees.
- Recommendation first — write the one-sentence recommendation before any analysis.
- Record one pass — capture one answer a week and listen for the skipped move.
- Assumptions labeled — every invented number gets a reason attached in the same breath.
- Limits stated — finish two answers a session with the "what I would check" line.
- Cold rerun — revisit one card from last week with no warm-up; spaced repeat makes the moves automatic.
- Course rubric checked — if participation or Q&A is graded, its wording outranks any generic advice, including this page.
Quiz: which move does each question need?
Five situations from the drill families. Wrong picks reveal the correct move, so the quiz teaches while it scores.
1. A cold call lands on you: "Is the winter ferry route viable?" Your first spoken move is to…
2. The bicycle-parking estimate is on the clock. The outline that earns the credit:
3. Pushback arrives: "We tried late opening in 2024 and it flopped." The strongest response:
4. The judgment question has thin data and a deadline. What earns the marks?
5. Ten seconds left in a 90-second answer. The best use of them:
Live answers are easy when the written case behind them is solid
Every good spoken answer points back at an analysis that exists on paper. If the written case study is the part running out of time, a deadline-bound expert writer can take it off your hands and return a draft you can defend line by line — exhibits and all.
Frequently asked questions
What are case study interview questions?
Prompts about a case that ask you to clarify, structure, estimate, judge, recommend or defend, fired at you live in a seminar, an exam or a hiring round. The setting changes the stakes, not the moves. Six families cover almost all of them, and each rewards a different first sentence — which is what the drill below trains.
How long should an answer to a case question take?
Roughly 60 to 90 seconds for a full answer and about 20 for a clarifying one. The clock matters less than the order: structure first, evidence second, decision last. An answer that runs long usually means the structure came late, so the listener held every fact before learning what it was for.
What if you do not know the answer to a case question?
Say the method instead of the number. Name the two or three things you would need, pick the most likely direction with the facts you do have, and state plainly what you would check. In seminars and hiring rounds alike, a structured partial answer beats a confident guess or a frozen silence.
Are seminar cold calls the same as consulting case interviews?
The question families are the same; the scoring is not. A professor is watching whether the class can see a method, so structure and fair-mindedness earn most of the credit. An interviewer is probing for a hiring signal, so arithmetic, follow-ups and composure weigh heavier. Prepare the moves once, then tune the depth to the room.
How do you practice case questions alone?
Drill one question type per day with a timer: read the prompt, answer out loud in 90 seconds, then compare against a model outline and note which move you skipped. Record one pass a week and listen back. Answering in your head hides the hesitations a listener would catch, so spoken practice is what transfers.
Did these drills sharpen your answers?