Oliver Wyman case interview: 4 practice cases with worked examples (2026)

The Oliver Wyman case interview is commonly candidate-led: the interviewer reads a short prompt, then expects you to set the structure, ask for the data you need and drive to a recommendation, often in around 30 to 40 minutes. Interviewers often act as collaborators, so think out loud and use their hints. Processes commonly run two rounds of case and fit interviews, with partners in the final round, and some offices add a written case or an online test. Financial services, insurance and risk appear more often than at most strategy firms, so ratios and segment economics matter.

Oliver Wyman was founded in the 1980s to advise financial institutions, and financial services and risk remain among its best-known practices. It now works across most sectors, including retail, transport, health and energy, so a case can come from anywhere. Its style shows even outside banking and insurance: interviewers like a ratio broken into drivers, an average cut by segment and a return compared with a hurdle. The fit portion is commonly a conversation about your experience and why Oliver Wyman; have true stories with numbers ready.

Two cases in our library are written in Oliver Wyman's style, both candidate-led and both with guided walkthroughs. We have added two banking cases written in other firms' styles, BCG and Accenture, because they test the same economics: what a decision costs across a whole book, and whether a business case survives realistic assumptions.

Start with Ashcombe: one ratio to rebuild and one segment to find. Fernpay adds capital and return on equity. Northstar is the hardest: the CEO asks a yes-or-no question, and the right answer depends on which customers you count.

The core skills are the same at every strategy firm: structure, maths, exhibit reading and a clear recommendation. The differences are in emphasis, and they vary by office and interviewer, so read this table as tendencies candidates commonly report, not rules.

The biggest adjustment is ownership. In a candidate-led case nobody tells you which exhibit comes next: you ask for it, and the quality of the request is part of the test. 'Can I see claims frequency by channel and driver age?' earns data; 'Is there any more information?' earns a pause. Our McKinsey and Kearney guides cover the interviewer-led style for contrast.

Most Oliver Wyman-style cases are built from the same seven kinds of question, and each can be drilled on its own.

Ashcombe Insurance: why the motor book lost money

The prompt. Ashcombe, a mid-size motor insurer, had a combined ratio of 96% two years ago. This year it is above 100%, so the motor book loses money on underwriting. Most new customers now arrive through price comparison websites. What went wrong, and how should it be fixed?

What clarifying questions earn you. The expense ratio is flat at 26%, so the movement is in claims. Across the market, claims inflation has pushed the cost of each claim up about 8% in two years, and competitors have raised prices about 10%. About 55% of Ashcombe's new business comes through comparison sites. The CEO wants the combined ratio below 97% within two years.

Two calculations. Claims cost per policy is frequency × severity: 7.6% × $5,400 ≈ $410 today, against 7.0% × $5,000 = $350 two years ago (a 70% loss ratio on $500). Today's loss ratio is $410 ÷ $533 ≈ 77%, so the combined ratio is 77% + 26% = 103%, up 7 points. Claims cost per policy has risen about 17% while premium rose 6.6%. On about $1.17bn of premium, 103% means an underwriting loss of about $35m.

Oliver WymanAshcombe Insurance's motor book slips into lossProfitability · Medium

Ashcombe Insurance's motor book slips into loss

Play the rest: find the segment behind the rise, put a dollar figure on it and decide the fix.

Interviewer

Our client, Ashcombe Insurance, is a mid-size motor insurer. Two years ago its combined ratio was 96%; this year it is above 100%, so the motor book is losing money on underwriting. Most of its new customers now arrive through price comparison websites. The CEO wants to know what has gone wrong and how to fix it. Where would you like to start?

Start the caseGuided, step by step

Fernpay: should a payments app start lending?

The prompt. Fernpay runs a payments app with 8m monthly active users, most aged 20 to 35. It has never lent money. Should it launch a credit card for its users, and if so, how?

What clarifying questions earn you. Fernpay holds an e-money licence but no lending licence, has no deposits and has $90m of cash. The market holds $80bn of card balances across 32m cards. The board wants at least a 15% return on equity and no damage to a brand built on young users' trust.

About half of users would pass affordability and credit checks, and the average balance is about $2,000. The first exhibit, a bar chart in the case, shows take-up at four comparable apps.

Two calculations. Average take-up is (14 + 17 + 23 + 26) ÷ 4 = 20%. So 8m users × 50% eligible × 20% = 800k cards, and 800k × $2,000 = $1.6bn of balances, 2% of the market's $80bn. The peer range gives 560k to 1.04m cards, or about $1.1bn to $2.1bn.

Oliver WymanFernpay weighs a credit cardMarket Entry · Medium

Fernpay weighs a credit card

Play the rest: choose a loss rate, build the return on equity and pick the launch route.

Interviewer

Our client, Fernpay, runs a payments app with 8m monthly active users, most of them aged 20 to 35. It has never lent money. The board is considering launching a credit card for its users and has asked us whether it should, and if so how. How would you approach this?

Start the caseGuided, step by step

These two are written in BCG and Accenture styles. Northstar is candidate-led, like an Oliver Wyman case; Fenmore is interviewer-led. Both turn on financial services economics that Oliver Wyman interviewers probe.

Northstar Bank: match the challenger's rate?

The prompt. A digital-only challenger has been taking deposits from Northstar, a mid-size UK retail bank: £1.66bn has left in 18 months, about 9% of the £18.4bn it holds today. The CEO asks whether to match the challenger's savings rate.

A fitting structure. Who is leaving and how much they hold; what matching costs across the whole book against what losing the deposits costs; responses other than price; then a decision segment by segment rather than for the book as a whole.

The exhibit's note adds that almost all of the £1.66bn left from instant access, and that the alternative to deposits is wholesale funding at 4.6%. Two calculations. Instant access is £11.4bn ÷ £18.4bn ≈ 62% of the book, and the rate gap is 4.1% - 2.2% = 1.9 points. A match would pay that 1.9 points on every pound of instant access, not just on the customers at risk.

BCGNorthstar Bank loses its depositsCompetitive Response · Hard

Northstar Bank loses its deposits

Play the rest: price the match and the outflow, profile the leavers and design the response.

Interviewer

Northstar is a mid-size retail bank. Over eighteen months a digital-only challenger has been taking deposit share in their core market, and Northstar has lost about 9% of retail deposit balances. The CEO is asking whether to match the challenger's savings rate. What do you think?

Start the caseGuided, step by step

Fenmore Bank: does a small-business app pay back?

The prompt. Fenmore Bank has 50,000 small-business customers and is losing them to digital banks at twice the rate of three years ago. The digital team wants $16m to build a mobile-first app. Does it pay back within the bank's four-year hurdle, and should Fenmore go ahead?

What clarifying questions earn you. Attrition is 12% a year, up from 6%, and most leavers cite the digital experience. Each customer brings about $2,000 of revenue a year, a $100m book. A few vendors also sell white-label small-business platforms.

A fitting structure. Adoption × value per active user, less build and run costs, year by year against the four-year hurdle. Then test the adoption assumption against peers, and compare building with buying.

Two calculations. In year 3 the plan has 60% × 50,000 = 30,000 active users, against 35% × 50,000 = 17,500 at the peer median, about 1.7 times as many. Over five years the plan adds up to 120,000 user-years (10,000 + 20,000 + 3 × 30,000) against 75,000 for the peers: 60% more.

AccentureFenmore Bank weighs a small-business appNew Product · Medium

Fenmore Bank weighs a small-business app

Play the rest: value an active user, test payback on both curves and compare building with a partner platform.

Interviewer

Fenmore Bank is a regional bank with 50,000 small-business customers, and it is losing them to digital banks at twice the rate it did three years ago. The digital team wants to build a mobile-first app for small businesses at a cost of $16m. The CEO asks whether it pays back within the bank's four-year hurdle, and whether Fenmore should go ahead. How would you approach it?

Start the caseGuided, step by step

Cases can come from any sector, but the firm's financial services roots show in two ways: the vocabulary, and the habit of judging a business by the return on the capital it ties up. You don't need industry knowledge for a generalist role, but you should be comfortable when a prompt uses these terms.

Written cases and online tests. Some candidates report a written or presentation-style case in later rounds: a pack of data, time to prepare, then a discussion. Some offices also use an online test before interviews. Whether you face either varies by office and year, so ask your recruiter. In a written case, find the two or three numbers that matter, put the answer first and show the arithmetic you rely on.

Before the interview, you should be able to say yes to each of these without hesitating.

• I can break combined ratio, margin or return on equity into drivers without notes.
• When an average moves, I check the segments before I conclude behaviour has changed.
• After my structure, I say where I'd start and ask for specific data.
• I size a lending book in balances as well as accounts.
• I turn a return into a breakeven: how far can the key assumption move before the answer flips?
• I close with an answer, the numbers behind it, the trade-off I am accepting and specific risks.

Play each case twice: first live against the clock, leading it as you would in the interview, then with the guided walkthrough, comparing each answer with the model. The gaps you find on the second pass are your study plan.

1. Week 1: Ashcombe, live then guided. Drill ratios and mix until a segment table takes you two minutes.
2. Week 2: Fernpay and Fenmore. Drill returns and payback; size every book in balances as well as customers.
3. Week 3: Northstar, then replay your weakest case and practise 60-second recommendations out loud.