Case Interview Examples: Official Cases Plus Three Worked Cases

What makes a good case interview example?

A good case interview example gives you four things: a clear client question, data that the interviewer reveals only when you ask for it, math that has one right answer, and a model close you can measure yourself against. A prompt alone trains structuring. A full example trains the whole 30 minutes, including the parts candidates fail most: exhibit reading, arithmetic under pressure and the final recommendation.

This page does two jobs. It lists the official practice cases each firm publishes, checked on 19 October 2026, and it adds three original cases, each written for a different interview format. For opening prompts only, use our list of case interview questions. For the full interview process, see the complete case interview guide.

Official practice cases from each firm

Official cases show how each firm wants a case to flow, so run them before anything else. The table records what each firm's own page showed on 19 October 2026. Links and lists change, so confirm on the firm's site before you rely on a count.

Three notes on the table. McKinsey's page says the set contains logic and suggested answers and that no single solution is correct. Bain's page also links three beginner and intermediate cases hosted by a third-party prep provider. The Kearney case comes from a September 2026 prep guide citing Kearney's site; we could not load Kearney's page to confirm it.

BCG's global case preparation page did not load on 19 October 2026. Its regional pages describe the format (cases based on real projects, clarifying questions expected, no single right answer) without naming a practice case. See our BCG case interview guide for its online screens.

Which official cases come with model answers?

McKinsey's five cases, Bain's Coffee Shop Co. and PwC's two scenarios all show suggested answers on the page. PwC goes further than most: it gives a sample answer for undergraduate candidates and another for MBA-level candidates, so you can grade yourself at the right bar. Deloitte keeps its practice inside an interactive tool and recommends a five-step approach that ends with next steps and expected impact.

One tool sits outside the table. McKinsey launched a free AI practice tool for quantitative case questions in April 2026 for entry-level candidates invited to interview, and about 10,000 people used it in its first month (reported by Business Insider). It drills the math sections; it does not replace a live partner. Firm-specific advice sits in our McKinsey case interview guide.

Worked example 1: an interviewer-led case with an exhibit (McKinsey style)

Prompt (fictional): Delta Bikes, a Dutch e-bike maker, grew revenue from €216M to €235M in two years, yet operating profit fell from €19.8M to €9.0M. The CEO asks why and what to do.

Interviewer script. Question 1: "How would you structure this?" Reveal on request: volume rose from 90,000 to 100,000 bikes. Question 2: show the exhibit and ask what drives the decline. Question 3: "What would you do about it, and what is it worth?"

Exhibit: economics per bike (€)

Math. Profit per bike fell by €130: €50 from price, €40 from battery and motor, €40 from warranty and logistics. At 100,000 bikes that is €5.0M, €4.0M and €4.0M, or €13.0M in total. Growth offset part of it: 10,000 extra bikes at the old €220 margin added €2.2M. Check: 19.8 - 13.0 + 2.2 = 9.0.

Data on request: warranty claims doubled on battery packs from a new supplier introduced in 2025.

Model close (60 seconds): "Profit fell €10.8M mainly because each bike earns €130 less: deeper dealer discounts, costlier batteries and a jump in warranty claims tied to one battery supplier. I recommend three moves: hold the supplier to its quality terms or switch, cap dealer discounts at 2024 levels, and renegotiate battery prices on higher volume. Halving the warranty and discount increases alone recovers about €4.5M. The main risk is dealer pushback, so I would start with the ten largest dealers."

Worked example 2: a candidate-led case with deal math (Bain style)

Prompt (fictional): A private equity fund is invited to bid for Rosmarin, a German drugstore chain with 400 stores. The seller asks €900M. The partner wants to know whether to bid, and at what price.

Interviewer script. In a candidate-led case you drive. The interviewer answers only what you ask, so the data below stays hidden until you request it.

Data on request: revenue €2.0B; EBITDA €100M (5% margin); comparable drugstore deals closed at about 8x EBITDA; private label is 20% of sales and earns a gross margin 10 points higher than branded goods; 40 small stores each lose €0.25M EBITDA a year; the improvement plan costs €30M once.

Math.

1. Asking multiple: 900 / 100 = 9.0x, one turn above peers. Standalone value at 8x: €800M.
2. Private label from 20% to 25% of sales: 5% of €2.0B = €100M of sales, times 10 points = €10M EBITDA.
3. Closing 40 loss-making stores: 40 x €0.25M = €10M EBITDA.
4. Plan value: €20M x 8 = €160M, minus €30M one-off cost = €130M net.
5. Payback on the €30M: 30 / 20 = 1.5 years.
6. At €900M the fund pays 800 + 100, handing the seller 100 / 130 = 77% of the plan's net value.

Model close: "Bid, but below the ask. Rosmarin is worth about €800M as it stands and about €930M if our plan works fully. Paying €900M gives the seller three quarters of the upside we would create. I recommend an opening bid near €840M and a walk-away price of €880M. The plan pays back in 18 months, and the main risk is the private label target, so I would test it in 50 stores during due diligence."

The Bain case interview guide covers how Bain uses deal cases like this one.

Worked example 3: an open-ended creative case (BCG style)

Prompt (fictional): Brouwerij Vandersteen, a family brewery in Belgium, sells 8,000 hectoliters of non-alcoholic beer a year. The CEO wants to double that in three years and asks for ideas.

Interviewer script. Ask for a brainstorm first, then for one sizing step. Reveal nothing else; this case tests idea quality and judgment.

A structured brainstorm. Group ideas by where to play and how to win:

• Where to play: supermarkets, cafés and restaurants, sports clubs and events, workplaces, export to the Netherlands and France.
• How to win: product (new flavors, cans, draught version), price (premium pricing at parity with regular beer), brand (separate label aimed at drivers and athletes), distribution (draught taps shared with the regular range).

Sizing step (practice assumptions): 9 million adults, 30% of whom drink non-alcoholic beer, at 20 bottles of 25 cl a year. That gives 9M x 0.3 x 20 = 54M bottles, or 54M x 0.25 l = 13.5M liters, which is 135,000 hectoliters. Vandersteen's 8,000 hl is about 6%, so doubling means reaching about 12% of this assumed market.

Model close: "Doubling is plausible because the brewery holds about 6% of a market of roughly 135,000 hectoliters. I would prioritize a draught version for cafés already pouring Vandersteen and a can format for sports venues, since both use existing customers. Next I would test willingness to pay at parity pricing in 20 cafés."

How to use these cases alone and with a partner

Use the official cases solo first: read the prompt, set a timer, write your structure, then compare with the suggested answer. Use the three worked cases with a partner, who reads the interviewer script and reveals data only on request. Rotate formats so each week covers one interviewer-led, one candidate-led and one creative case. Before you structure, sketch a first cut using our guide to case interview frameworks.

Cracking the Case adds three fully worked candidate cases (30+ pages) built the same way, with interviewer prompts and model answers.

Club casebooks: useful, but check the year

Business school consulting clubs publish casebooks with dozens of cases each, and they remain good drill material for structuring and math. Many circulating copies date from 2001-2018, before online screens, AI rules and the current fit-interview formats. Use them for repetition, and use the official cases above to calibrate what today's interviewers expect.