CaseClimbAsk ten people how many cases you need to practice before a real consulting interview and you'll get ten different numbers, most of them guesses. The honest answer is a range: most candidates need somewhere between 30 and 40 full, timed cases of deliberate practice before they walk in ready. Below that, the gaps tend to be structural. You don't yet have a repeatable process for turning an ambiguous prompt into a structure, and no amount of luck with the case draw covers for that. "I got unlucky with a hard case" is the story candidates tell afterwards; the missing process is the cause.
Why 30-40, and not 5 or 100
Five cases is enough to learn what a case interview looks like: the format, the pacing, the kind of math the interviewer will hand you on the spot. It is not enough to make any of it automatic. Under real interview pressure, anything you haven't practiced into a habit tends to disappear. That is why candidates who have done a handful of cases often freeze on the exact skill they would tell you they are good at.
At the other end, there is a ceiling past which more reps stop paying off. Once you can build a tailored, MECE structure in under two minutes on most prompts, get the math right under time pressure, and land a recommendation with a risk attached, case number 60 teaches you very little that case number 35 didn't. Past that point, book a live mock interview where someone can push back on you the way a real interviewer would. You will get more from that hour than from another solo case.
Reps across variety beat the same case forty times
The number matters less than what's inside it. Thirty cases that are all profitability diagnostics will make you very good at profitability diagnostics and leave you exposed the moment you draw a market-entry or pricing case. A well-built prep plan spreads practice across case types (profitability, growth, market entry, pricing, M&A, operations) so that you find the gap in a type you have not touched during practice, rather than in the interview itself.
The same logic applies to skills as well as case types. A case interview is five skills wearing a trenchcoat: structuring, quantitative analysis, business judgement, synthesis and recommendation, and communication. You can be strong on four of them and weak on one across attempt after attempt. Say you build clean frameworks and get the numbers right, but your final recommendations skip the risk and the next step more often than not. Volume alone won't fix that. You have to know which of the five is lagging and practice that one on its own.
How to tell where you stand
- You can build a tailored structure for a new prompt in under two minutes without a memorized template.
- Your math is both accurate and fast enough that the interviewer is not waiting on you.
- You reach a recommendation before the interviewer asks for one, and it comes with a quantified risk and a next step.
- Your performance doesn't collapse the moment an interviewer pushes back on your first hypothesis.
- You have worked through at least four or five different case types, including the ones you would rather avoid.
If most of those aren't true yet, put in more deliberate practice before you worry about confidence. If they are true, your next step is a live mock interview rather than case number fifty.
