How to practise case interviews on your own: a two-week plan

The usual advice is to do 30 or 40 cases with a partner. That's fine if you have a partner with free evenings and the patience to read a case aloud properly. Most people don't. They have a friend who can manage two sessions a week and a stack of casebooks they've read but never spoken.

Reading cases is not practising cases. You can follow a perfect answer on the page and still freeze when someone says “so where would you start?” and waits. The skill being tested is thinking out loud under mild pressure, and the only way to build it is to do exactly that, on your own if you have to.

Here is the plan I'd give a friend with two weeks and no reliable partner. It assumes about 90 minutes a day on weekdays and a bit more at the weekend.

Talk out loud every single session

This is the rule that makes everything else work. If you're doing a case alone, say the whole thing aloud, including the clarifying questions you'd ask. Record it on your phone. It will feel ridiculous for about two days and then it won't.

Recording matters because you can't judge your own structure while you're producing it. Played back at 1.5x speed, the problems are obvious: the 20-second silence before your framework, the “um, so, basically” that opens every sentence, the moment you said “I'd look at costs” and never came back to it.

Week one: build the pieces

Week one is about components. You're not trying to do full cases well yet. You're trying to make each part automatic so that in week two your attention goes on the case, not on remembering what comes next.

Days 1–2: openings and clarifying questions

Take ten case prompts from any casebook. For each one, do only the first two minutes: repeat the prompt back in one sentence, ask two or three clarifying questions and state the objective. Stop there. Ten openings take about 30 minutes and you'll notice your clarifying questions getting sharper by the sixth.

Good clarifying questions narrow the problem. “What does success look like for the client?” is useful. “Can you tell me more about the company?” is not, because it hands the work back to the interviewer.

Days 3–4: structures

Same ten prompts. Now take 60 seconds of silence, write a structure, then present it aloud in under 90 seconds. Aim for three or four branches, each with two or three specific questions underneath. Then compare against the casebook answer and write down one thing you missed.

If you're starting with profitability (most people should), the breakdown in how to structure a profitability case is a sensible base to tailor from.

Day 5: maths

Spend a whole session on mental arithmetic: percentages, growth rates, break-evens, dividing large numbers. Fifteen minutes a day after this, every day, until the interview. Our post on case interview maths without a calculator has the shortcuts worth drilling.

Days 6–7: exhibits and the “so what”

Pull every chart and table you can find from casebooks. For each, give yourself 30 seconds, then say out loud: what the exhibit shows, the one number that matters and what it means for the client. That third part is where most candidates stop short. “Costs went up 12%” is a description. “Costs went up 12%, nearly all of it fuel, so this is a cost problem and not a demand one” is an insight.

Week two: full cases, then fix the weak stage

Now you string it together. Do one full case a day, timed, recorded, start to finish. Then review the recording with a simple scorecard.

After three full cases you'll see a pattern. Most people have one stage that drags down everything else. For many it's synthesis: they do a decent analysis and then ramble through it in order instead of opening with the answer. Spend days 11 and 12 drilling only that stage on old cases you've already done.

Days 13 and 14: two more full cases, ideally with a real person if you can get one, and a light final day. Don't cram the night before. One clean case and an early night beats three tired ones.

What practising alone can't give you

The weakness of solo practice is that you know the case is coming. You've read the prompt, so you also know what data exists. In a real interview the interviewer holds information back and only shares the exhibit when you ask the question that earns it. They also push back: “Why would you look at that first?” or “I'm not sure that's right, check your number.”

You can get some of this from a friend, but they'll often give you the data early because they're being nice. If you want an interviewer who keeps the exhibit back until you ask for it and questions your maths, that's what RoundReady's consulting track is built to do. Either way, try to do at least a few cases where you genuinely don't know what's coming.

Common mistakes in solo prep

1. Doing more cases instead of better ones. Ten reviewed cases teach more than thirty unreviewed ones.
2. Memorising frameworks. Interviewers have heard “the 4Ps” a thousand times. They want branches that fit this client.
3. Skipping the maths drills because they're boring. A wrong number halfway through a case is very hard to recover from.
4. Never timing yourself. Most first-round cases run about 25 to 30 minutes. If your structure alone takes five, you'll run out of road.
5. Only practising one case type. Do at least a market entry, a pricing case and a market sizing as well as profitability.

What to say when you get stuck

You will get stuck, in practice and probably in the real thing. The fix is to narrate. Interviewers can work with a candidate who says what they're thinking. They can't work with silence.

“I've ruled out volume, since passenger numbers are flat. So it's either price or mix. Could I check whether the average fare has moved?”

That sentence does three things. It shows what you've already eliminated, it keeps your structure visible and it asks for a specific piece of data. Practise saying lines like that until they come out without thinking, and being stuck stops being a disaster and becomes a normal part of the case.