How to Close a Case Interview: The 60-Second Recommendation

What should a case interview recommendation contain?

A strong recommendation has four parts in a fixed order: the answer, the reasons, the risks and the next step. The answer comes first because the interviewer is playing a busy client who may stop listening after one sentence. The reasons carry the numbers you produced during the case, so the close also proves you remember and understood your own analysis.

This order is the Pyramid Principle applied to speech: conclusion on top, supporting arguments below, each argument backed by evidence. The Minto Pyramid Principle article covers the theory. This guide turns it into a script you can deliver under a timer.

Official firm guidance points the same way. BCG lists clear communication and logical reasoning among the qualities its interviewers look for. Bain advises candidates to highlight key numbers and insights. Strategy& tells candidates to take a stand at the end of the case even when data and time feel short.

The 60-second script, second by second

Sixty seconds is about 150 spoken words. Split them like this.

Three rules make the script work. First, the reasons must be mutually exclusive: market, economics and feasibility are separate; "demand is high" and "customers want it" are the same reason twice (the MECE guide shows how to test this). Second, every reason carries a number. Third, the risk section names a risk you can mitigate together with the mitigation, which shows judgment.

Mid-case synthesis: when and how to summarize

The final close is easy when you have summarized along the way. A short synthesis belongs at three moments: after you finish an exhibit or calculation, before you switch to a new branch of your structure, and whenever the interviewer asks where you stand.

Each mid-case summary has three beats: what we now know, what it means for the client's question, and what to look at next. For example: "Rental margins are 45%, well above the 30% the client earns on tool sales. That makes the economics attractive, so the open question is whether demand is large enough. I would like to look at customer data next."

An interviewer rarely needs a long summary at these moments. What the interviewer looks for is evidence that the candidate is steering toward an answer, and a 15-second check-in provides it.

Worked example: a weak close rewritten

The case. Brugman, a fictional Dutch DIY retailer with 140 stores, asks whether it should launch a tool-rental service. During the case the candidate found the following: rental gross margin of 45% versus 30% on tool sales; a customer survey in which one in three shoppers said they bought a tool they used once; an estimated EUR 180,000 of rental revenue per store per year (25 rentals a day x EUR 20 x 360 days); a tool fleet investment of about EUR 100,000 per store; and a risk that rentals replace some sales.

The weak close (about 60 seconds, no answer until the end):

"So, we looked at a lot of things. First the market, which seems decent, then the customers, where the survey was quite interesting, and then the costs. There are some positives, like the margins, but also some negatives, like cannibalization and damaged tools. Overall I think it could make sense for them to consider rental, maybe in some stores, depending on more data."

The problems: no answer in the first sentence, no numbers, a list of topics where reasons should be, and a conclusion hedged into nothing.

The rewrite (about 150 words, 60 seconds):

"Brugman should launch tool rental, starting with a six-month pilot in 20 stores. [Answer, 10 s]

Three reasons. First, the economics: rental earns a 45% margin against 30% on sales, so each store adds about EUR 80,000 of gross profit on EUR 180,000 of rental revenue. Second, demand: one in three shoppers bought a tool they used once, and those customers are the natural rental market. Third, payback: at about EUR 100,000 of fleet per store, the investment pays back in a little over a year of gross profit. [Reasons, 35 s]

The main risk is that rentals replace sales of the same tools. The pilot should measure this directly by comparing sales in pilot and control stores. [Risk, 10 s]

As a next step, I would select the 20 pilot stores this month. [Next step, 5 s]"

Check the arithmetic before you say it: 45% of EUR 180,000 is EUR 81,000, rounded to EUR 80,000, and EUR 100,000 / EUR 81,000 is about 1.2 years.

Recording the Casey video recommendation

BCG's Casey online case ends with a webcam recommendation. MyConsultingCoach reports a limit of about 60 seconds with a strict cutoff, and candidates cannot return to earlier questions. That cutoff is the strongest argument for answer-first: if the recording stops at second 50, a candidate who opened with the answer has still delivered it.

Four habits help on camera:

1. Write four keywords before you press record: answer, reason one, reason two, risk.
2. Pull numbers from your earlier answers. Casey's questions build on each other, so your notes already contain the evidence.
3. Look at the lens, with your keywords placed just below it.
4. Rehearse with a 55-second timer so you finish with margin.

More on BCG's process is in the BCG case interview guide, and how Casey compares with other screens is in the consulting online assessment overview.

Handling "the CEO walks in" and other closing prompts

Interviewers often trigger the close with a scene: the CEO steps into the room, you share an elevator with the client, or you have 30 seconds before a board meeting. All of these ask for the same script, compressed. Drop to one reason with its number, keep the risk to one clause, and keep the next step.

When the prompt arrives mid-case, before your analysis is complete, give a hypothesis-based answer: "Based on what we have so far, I would recommend X, because of Y. The one thing I would still verify is Z, and if Z is negative, I would change my answer." This shows you can commit and still stay honest about uncertainty.

The same applies at the end of an AI-assisted exercise. Reports on McKinsey's pilot with its Lilli tool, which McKinsey has not commented on, say candidates are assessed partly on communicating a clear, structured answer after working with the tool. Add one sentence on which figures you checked yourself, then deliver the script as usual.

Stephan's note: The close is the cheapest part of the case to improve. Record ten closes on your phone, play them back, and count the seconds until the answer arrives. If it takes longer than ten, rewrite the first sentence and record again.

Avoiding the classic closing errors, such as summarizing the process instead of the answer, is covered in case interview mistakes, and the full interview flow is in the complete case interview guide. My book Cracking the Case includes three fully worked candidate cases, each ending with a recommendation you can use as a model.