Case Interview Synthesis: Answer First, Then Next Step

The one published case interview sequence with a step after the call

Deloitte's tips page at deloitte.com publishes a five-step approach with these exact titles: "Understand the issue", "Identify the underlying assumptions", "Summarize specific issues and findings", "State your recommendations", "Outline next steps and expected results/impacts".

Count the last two. Recommendation is step four. Step five is separate, and it asks for two things: what happens next, and what you expect it to produce.

Now compare the other published sequences. L.E.K.'s seven steps end at "Come up with an answer", glossed as committing to a course of action. Oliver Wyman's six method steps end at make a recommendation. Accenture's workbook sequence ends at Summarize findings. Kearney's four tips end on making the experience your own. Of the seven sequences in our corpus, six end at the recommendation.

The same Deloitte page also sets the standard the close is judged against: "When working through a case or scenario interview, you are not expected to provide the "right" answer, but rather to clearly convey your logic and thought process."

That combination is the argument for a next step. If logic is what is assessed, a recommendation with no test attached is logic that stops one move early. A next step says how you would find out whether you are right.

Never end a case interview on a number, and what to end on instead

A Kearney business analyst in the Chicago office published a line on her employer's careers blog that sits at exactly the same joint. Read through a text proxy at kearney.com, she writes: "When you're working through the quantitative part of the case, you should never end on a number."

That is a named consultant describing her own practice, published by her employer. It is not a firm policy statement, and we label it that way.

The instruction is narrow and useful. A number is an output. What the interviewer is waiting for is the sentence that converts it into a decision, and there are only three shapes that sentence takes.

1. It clears a bar. The payback is 2.4 years against a 3-year hurdle, so this passes.
2. It fails a bar. The payback is 8 years and the equipment lasts 6, so this cannot work at these assumptions.
3. It brackets a bar. The payback is 3.1 years against a 3-year hurdle, so it turns on the one assumption I flagged.

Every one of those names the bar. The bar is the thing candidates most often leave out, and without it a number has nothing to be compared to. If the interviewer has not given you a hurdle, ask for it, or state the one you are assuming.

A case interview synthesis written out, with its payback

Here is a complete close, on a case with stated assumptions, so the arithmetic is auditable.

The situation. A nine-store regional chain earns 1.9 million dollars of operating profit. Two stores lose money: one loses 180,000 dollars a year, the other 240,000. Exiting each lease costs 500,000 dollars. The client applies a three-year payback hurdle.

The close, as you would say it.

Close both loss-making stores. That lifts operating profit from 1.9 million to 2.32 million dollars, which is 22.1 percent, and the 1 million dollars of lease exit pays back in 2.38 years against your three-year hurdle.

The evidence is two lines. The two stores together lose 420,000 dollars a year, so closing them removes that loss entirely. The exit cost is 500,000 dollars each, so 1 million dollars in total, and 1 million divided by 420,000 is 2.38 years.

The risk is that some of their revenue walks rather than disappears. The two stores turn over 3.0 and 3.4 million dollars, so 6.4 million. If 15 percent transfers to nearby stores at a 30 percent contribution margin, that adds 288,000 dollars of contribution and the payback shortens to 1.41 years.

Next step is a two-week trade area study on the four stores nearest the two closures, to measure the transfer rate. If it comes in above 15 percent the case gets stronger, and if it comes in near zero the payback is the 2.38 years I quoted.

Every figure recomputed by hand on 30 September 2026. 180,000 plus 240,000 is 420,000. 1.9 million plus 0.42 million is 2.32 million. 0.42 over 1.9 is 22.1 percent, rounded from 22.105. Two leases at 500,000 is 1 million. 1,000,000 over 420,000 is 2.381, so 2.38 years. Revenue 3.0 plus 3.4 is 6.4 million. 15 percent of 6.4 million is 960,000, and 30 percent of that is 288,000. 420,000 plus 288,000 is 708,000, and 1,000,000 over 708,000 is 1.412, so 1.41 years.

The risk sentence, and why it shortens the payback here

Most candidates treat the risk section as a hedge. In the close above it does the opposite, and that is worth understanding because it is common.

The risk named is revenue transfer. Transfer is a risk to the size of the saving, because the two stores' sales do not all vanish. But every dollar that transfers arrives at a store the client still owns, and it arrives with contribution margin attached. So pricing the risk improves the recommendation rather than weakening it.

The honest version says both directions in one breath. Transfer makes my saving figure wrong, and the error runs in my favor, so 2.38 years is the conservative reading.

Two firms publish evidence that this material belongs in the close. Bain's coffee shop case separates a good answer from a better one, and writes that the better answer "would take into account how long it would take for her to recoup the cost to open the store and other considerations before proceeding", from bain.com. Deloitte's LeadAuto case at usrecruiting.deloitte.com separates "A good answer may include" from "In addition to the above, a great answer may also include", and the upper tier adds alternatives, a multi-year commitment, a joint venture and expansion beyond the current region.

Two firms, no connection, the same upper tier: alternatives, risk, horizon, next step. Two instances are two instances rather than a market pattern, and that is how we report it. What is fair to say is that both firms put this material above the line.

The four parts of a case interview synthesis, in sixty seconds

The close has a fixed order and the order is doing work.

That totals 54 seconds, which is why sixty is a workable cap. The number is ours; no firm publishes a target length for a synthesis.

The call goes first because everything after it is support, and a listener who does not know the destination cannot tell support from digression. Oliver Wyman's sixth method step is make a recommendation, at oliverwyman.com, and its published criteria include communicating ideas. A recommendation buried at the end of a narrative has been communicated badly even when it is correct.

The risk goes third rather than last because the next step usually tests the risk. Putting them in that order lets the last sentence follow from the one before it.

One failure mode is worth naming. A close that opens with I looked at three things has not started yet. The three things are the evidence, and they belong after the call.

Answer first carries 35 of 100

The weights below are ours, taken from the grading code in src/lib/grading. No consulting firm publishes a marking scheme, which is a constant across every firm file behind this site.

Drilling the close on its own

The synthesis is the most drillable part of a case after the opening, because you can practice it on a finished case you have already solved.

1. Take a case you solved last week and close it again in sixty seconds. Timed, out loud, from the call.
2. Write the four parts as four lines before you speak. Call, two figures, risk with its direction, next step with a timeframe.
3. Run the two-tier test. Does your close contain an alternative, a risk, a horizon and a next step? Missing all four is the lower tier of two firms' own published answers.
4. Name the bar every time you say a number. Payback against a hurdle, share against a target, margin against last year. No bar, no meaning.
5. Say which way the risk runs. Conservative or optimistic. A risk with no direction is a disclaimer.
6. Give the next step a timeframe. Two weeks, one quarter, before the board meeting. A test with no clock is a wish.

Our synthesis drill scores Answer first, Evidence, Risk and Next step as separate lines and names the first correction, and the breakeven calculator will check a payback while you build the close. Both are free and neither needs an account. Neither knows whether your recommendation is the right one for that client, which is the judgment the firm is hiring for.

For the arithmetic habits that feed a close, see case interview math tips. For the top-down discipline behind answer-first delivery, see the pyramid principle. For the opening at the other end of the case, see the case interview opening statement.