The 10 Most Common Case Interview Mistakes I Saw as an Interviewer

The most common case interview mistakes I saw as a McKinsey interviewer were rarely calculation errors. They were process mistakes: skipping clarifying questions, forcing a memorized framework onto a case it did not fit, building structure silently instead of saying it out loud, freezing when a number came out messy, and running out of time before delivering an actual recommendation. I sat across the table for over a hundred interviews, and these ten patterns showed up again and again, usually costing the candidate the case before the analysis even started.

Why these mistakes are so consistent

Case interviews test a narrow set of skills on purpose: structured thinking, quantitative reasoning, business judgment, and communication under pressure. Because the skill set is narrow, the ways candidates fail it are narrow too. After enough interviews, you stop seeing individual candidates fumbling in unique ways and start seeing the same handful of patterns wearing different clothes. None of the ten below require more case knowledge to fix. They require noticing you are doing them.

The 10 mistakes, ranked by how often I saw them

Every one of these is fixable in a weekend of deliberate practice. None require more case knowledge. They require catching yourself in the moment, which is exactly why I want to walk through the worst offenders in more detail.

Mistake 1: Diving in before you've restated the problem

A candidate hears the prompt, nods, and starts talking about revenue drivers within ten seconds. It feels efficient. It reads as reckless. The interviewer just handed you a set of facts, a client, and an objective, and your first job is proving you absorbed all three correctly, not proving you can talk fast. Restate the client, play the numbers back accurately, and state the objective in your own words before you do anything else. This single habit, done consistently, separates candidates who look like they are already on the job from candidates who look like they are still catching up. If you have never seen this done well end to end, a full sample case interview walks through it line by line.

Mistake 3: Forcing a memorized framework onto the case

This is the one I could spot in about ninety seconds, every time. A candidate memorized four or five named frameworks before the interview, hears a prompt, and starts jamming it into whichever framework feels closest, whether or not the boxes actually fit the client's problem. The tree ends up generic. Half the branches don't apply, and the candidate defends them anyway because that's what the framework said. The fix is to stop memorizing shapes and start building from an equation, profit equals revenue minus costs, revenue equals price times volume, market size equals population times adoption times price, then tailor the branches under that equation to the specific facts you were just given. I go through this build process in detail in how to structure a case interview, and the difference between a memorized framework and a tailored tree is covered fully in the case interview frameworks guide.

Mistake 4: Building structure silently

Some candidates do build a genuinely good tree, then say almost nothing while they draw it, and only speak again once they've picked a starting branch. The interviewer cannot grade thinking they cannot see. Consulting is a communication job as much as an analytical one, and a partner briefing a client does not hand over a silent diagram, they walk the room through the logic. Once you've used your structuring pause, signpost what you built: name the branches, say why each one matters, then propose where you want to start and why. That thirty seconds of narration is often worth more to your score than the tree itself.

Mistake 6: Arithmetic errors that go uncaught

Everyone makes a math slip under pressure, that part is forgivable and expected. What isn't forgivable is not noticing. If your revenue calculation implies the client sells more units than exist in the entire market, or a cost line comes out negative, or a percentage change is off by an order of magnitude, that should trip an alarm before you say the number out loud. Pause, sanity-check the output against the scale of the business, and if something looks wrong, say so and rework it. Interviewers read a caught error as good instincts. They read an uncaught error, especially a wildly implausible one, as a candidate who doesn't actually understand what the numbers represent.

Mistake 7: Narrating to fill silence

Silence feels dangerous in an interview, so candidates fill it, thinking out loud in fragments, half-sentences, and false starts. It feels like transparency. It usually reads as disorganization. The better move, once you're past your one legitimate structuring pause, is to think for a few seconds, form a complete thought, and then say it in one clean sentence. A consultant on a client call doesn't ramble through their reasoning live, they present conclusions and defend them when asked. Practicing this specific habit, pausing before speaking instead of speaking to think, is one of the fastest score improvements available to a candidate who already knows the content.

Mistake 9 and 10: No recommendation, or a recommendation with nothing behind it

Some candidates run a technically sound case and then just stop, waiting for the interviewer to ask "so what's your recommendation." Others give a recommendation but it's a bare sentence, no tradeoff, no risk, no next step. Both read as a candidate who did the math but never actually thought like a consultant. A client does not pay for analysis, they pay for a decision they can act on. Track your logic through the whole case so that when the moment for synthesis arrives, whether the interviewer asks or you volunteer it, you can state a recommendation, the one or two reasons behind it, the biggest risk to watch for, and what you'd check next. That closing structure is what makes an answer sound like a recommendation instead of a summary.

What ties all ten together

Look back at the list and none of these are knowledge gaps. Nobody fails a case interview because they didn't know that profit equals revenue minus costs. They fail because they skip a clarifying question they knew to ask, or narrate instead of think, or stop one sentence short of a real recommendation, under the pressure of being watched. The content of case prep, the frameworks, the math, the industry knowledge, is genuinely learnable in a few weeks. The habits above are learnable too, but only through repetition under some version of real pressure, which is why timed practice with feedback beats reading about cases every time.

The bottom line

Nine times out of ten, the case interview mistakes that cost candidates an offer happened before the real analysis started: a skipped clarifying question, a framework forced onto the wrong problem, a tree built without ever being said out loud. Fix the process and the content problems mostly take care of themselves. If you want to see the whole process done correctly from the opening line to the final recommendation, the beginner's guide to case interviews is the place to start before you drill the ten mistakes above out of your own habits.

Go deeper

Catching these ten mistakes in yourself is hard to do alone, which is exactly what the CaseMap method and the Interview Dance in the full course are built to drill out through structured, repeated practice.

Get the free resume module →

Get the complete Cut to the Case course →

One payment, lifetime access, 30-day guarantee.