How case interviews are scored: what interviewers are really marking

Candidates often walk out of a case interview thinking they did well because they “got the answer”. Then they don't get through. The reason is usually simple: the answer was never the main thing being scored.

Each firm has its own form and its own wording, but most strategy firms assess some version of the same four areas. Once you know what they are, a lot of standard case advice suddenly makes sense.

1. Structure: can you break a messy problem into pieces?

This is the first thing an interviewer forms a view on, often within the opening few minutes. They want to see a breakdown that covers the problem without overlapping and that's tailored to this client rather than recited from a book.

The giveaway of a memorised framework is that it doesn't fit. If the case is about a ferry operator losing passengers and your structure has a box for “brand positioning” before you've looked at timetables or pricing, the interviewer notices. A slightly rough structure that is clearly yours scores better than a polished one that isn't.

2. Maths: accurate and explained

Arithmetic errors hurt, but not always in the way people expect. One slip you catch yourself is barely a mark against you. A wrong number you build three more steps on, without noticing it looks odd, is a serious one. What's being tested is whether you can be trusted with numbers in front of a client.

Interviewers also mark how you get there. Saying the set-up first (“profit is revenue times margin, so I'll work out margin”) shows your logic even if the sum goes wrong. Silent scribbling followed by a number shows nothing.

3. Synthesis: what does it all mean?

This is the area most candidates underrate. At the end of a case (and often in the middle of one) the interviewer wants you to pull the findings into a clear view. Not a recap of everything you did. A recommendation with reasons.

The strong version leads with the answer, quantifies it and says what to do next. That's the format a partner wants from a consultant at the end of a long day and it's the format interviewers are listening for.

4. Communication: could they put you in front of a client?

Underneath every other score is a quieter question. Would I be comfortable with this person in a client meeting next month? That covers how clearly you speak, whether you signpost (“I'll look at three areas, starting with costs”) and how you react when you're stuck or corrected.

Being corrected is a test in itself. Interviewers will sometimes challenge a reasonable point just to see what happens. Getting defensive costs you. So does folding instantly. The best response is to think about it, then either update your view or explain calmly why you'd keep it.

“Before you go further, can you tell me what you think is driving this so far?”

If you hear a line like that mid-case, it's a synthesis check. Give a one-sentence hypothesis, even a tentative one. “I don't know yet” is honest but scores nothing.

Why the right answer isn't enough

Many cases don't have a single right answer. Even when they do, the interviewer already knows it. What they don't know is how you think. A candidate who reaches the expected conclusion by luck or by forcing a framework gives them little to write down. One who reaches a slightly different conclusion through clear, well-explained steps gives them plenty.

None of this is secret. It's just easy to forget under pressure, when getting to a number feels like the whole job. Treat the number as one part of the case and the rest of the score tends to follow.