How to Structure Any Case Interview in 90 Seconds

Most candidates lose the case in the first three minutes, and they lose it the same way: they hear a prompt, recognise a category, and reach for a framework they memorised. The interviewer has seen that framework four times this week.

Structuring well is not about knowing more frameworks. It is about a repeatable method for turning an objective into branches you could actually collect data about. Here is that method.

Why memorised frameworks get flagged

The classic frameworks were written to organise strategy documents, not to solve a specific client's problem in twenty minutes. Recited verbatim, they signal that you stopped thinking as soon as you recognised the case type.

There is a simple test. If your four buckets would be identical for an airline and a software company, you have not structured. You have recalled.

Keep the classics as a private checklist. After you build your tree, scan them silently to see whether you have missed a whole area. That way they improve your exhaustiveness without flattening your tailoring.

The method

Five steps, in order. The whole thing fits in ninety seconds once it is automatic.

1. Restate the objective as a number

"Grow EBITDA by twenty million in two years" is a tree root. "Improve performance" is not.

If the prompt did not give you a number, ask for one. You need to know what success looks like and over what horizon before you can decide what matters.

2. Split on an equation, not a theme

Your first cut should be arithmetic. Revenue and cost. Volume and price. Acquisition and retention.

Equations make branches sizable. Themes do not. Compare these two openings for a declining profit case:

Profit is revenue minus cost. Revenue is volume times price. Cost splits into fixed and variable.

I would look at customers, competitors, costs and marketing.

The second sounds reasonable and is unusable. Marketing spend is a cost. Competitors act through volume and price. The same effect can be filed in three places, so nothing can be quantified cleanly.

The universal root worth memorising is the only formula you genuinely need:

Profit = (Volume x Price) - (Fixed Cost + Volume x Unit Variable Cost)

Nearly every profitability, pricing, growth and cost case hangs off that identity.

3. Split again on the client's reality

The second level is where you show you listened. Segments, channels, geographies, product lines: whatever the interviewer actually mentioned in the prompt.

This is the step that separates a tailored tree from a generic one, and it costs you nothing but attention.

4. Stop at testable leaves

Keep splitting until each leaf suggests a question you could ask. "What happened to our average price in the enterprise segment?" is a leaf. "Pricing strategy" is not.

If you cannot imagine the data that would sit under a branch, that branch is a label, not an analysis.

5. Prioritise out loud

This is the step most candidates skip, and it is the one that converts a static tree into a plan.

Name the branch you would start with and give a reason tied to the prompt:

Revenue grew twelve percent while profit fell thirty. That points at cost, so I would start with cost per unit rather than demand.

Six seconds. It changes how the whole structure reads.

MECE, properly understood

Mutually exclusive, collectively exhaustive. Candidates usually manage exhaustive and fail exclusive, producing structures where the same driver appears in three branches and the analysis goes in circles.

The test that catches overlap: for each branch, ask whether a finding there could also be true in another branch. If yes, your cut is wrong. Re-cut on something arithmetic rather than something thematic.

One nuance worth internalising. Perfect MECE is a tool, not a religion. A structure that is ninety-five percent exhaustive but genuinely tailored to this client beats a textbook-clean framework that could have been drawn before the prompt was read.

The sentence that does the most work

After presenting your tree, add one clause:

Given this is a subscription business, I would expect retention to matter more than acquisition, so I have split revenue that way.

That single sentence is the difference between a memorised framework and a tailored one. It tells the interviewer you chose this cut for this company, and it is almost impossible to fake.

What to practise

Structuring is scored on every exchange, not just the opening. Drill it the way it is used:

1. Take any prompt and build a tree in ninety seconds, out loud, against a timer.
2. Force yourself to cut arithmetically at the top level every time.
3. Finish with the prioritisation sentence. Every time, until it is automatic.
4. Then check it against the classic frameworks to see what you missed.

Ten repetitions of that loop will do more than reading another case book.