A recap
Walks back through the case in the order you did it.
- Opens with
- "So first we looked at..."
- Leaves out
- What to do on Monday
The case interview recommendation, how to close in a minute
End a case by stating your recommendation first, then two or three reasons with numbers, one main risk and one next step, in about a minute. Dartmouth's career center tells candidates to always give a concrete recommendation and name the barriers or risks to it. BCG says there isn't always a single right answer, so what counts is a clear position you can defend. We build Soreno, an AI voice interview practice app. Try a close on our free profitability case with no sign-up, then compare it with the worked close below.
Updated
- Length
- About a minute
- Ours, a BU‑hosted guide says under two
- Order
- Answer, reasons, risk, next step
- The answer always comes first
- Reasons
- Two or three, with numbers
- Taken from the case itself
- Worked case
- One airline route
- Invented by us, every number shown
How do you end a case interview?
Dartmouth's career center says to always end with a concrete recommendation and the barriers or risks to it. A guide by a Leland coach, republished on BU Questrom's career center blog, says to close on a top risk and one specific next step, and to practice the summary out loud in under two minutes. The same guide says to lead with your recommendation, not build up to it, and the timings below are ours. A synthesis is not a recap of what you did. It is what the client should do, and why.
A recommendation
Says what to do, then gives the proof.
- Opens with
- "You should launch the route"
- Ends on
- One risk and a next step
Worked example: Skyline Air (ours)4 parts
Skyline Air, an invented regional airline, asks whether to launch a new route. You have found 150 seats a flight, a 70% load factor, a $180 fare, 2 flights a day for 360 days, $17,000 of cost a flight and $3 million up front. All invented by Soreno for practice.
The answer, 10 seconds
Say the decision first
"Skyline should launch the route." No build-up. If the interviewer stopped you here, they would know where you landed.
The reasons, 30 seconds
Two reasons, each with a number
"First, it makes money: 720 flights a year at $18,900 of revenue a flight is $13.6 million, against $12.2 million of cost, so $1.4 million a year. Second, it pays back fast: the $3 million up front comes back in about 2.2 years."
The risk, 10 seconds
The one thing that would break it
"The main risk is demand. The route breaks even at a 63% load factor and we're assuming 70%, so there's little room."
The next step, 10 seconds
What happens on Monday
"Next, I'd start with one daily flight for six months and only add the second if load stays above 63%."
The same close, weak and strong (ours)2 versions
Both use the same numbers. Only one sounds like advice to a CEO.
Weak
A recap with no position
"So we looked at revenue and costs, and the revenue was $13.6 million and costs were $12.2 million. There are some risks around demand and competitors. It could work, but it depends, and I'd want more data before deciding."
Strong
A decision with its proof
"Skyline should launch the route. It earns $1.4 million a year and pays back $3 million in about 2.2 years. The main risk is demand, since it breaks even at a 63% load factor against the 70% we assume. So I'd start with one daily flight for six months, and add the second if load holds."
Pushback after you close (ours)3 questions
Most guides stop at the close. Here is what we'd ask next, and the answer we'd want.
On the risk
What if load only reaches 60%?
"Then each flight loses money, because break-even is 63%. That's why I'd start with one flight a day and set a date to check it, so we find out cheaply."
On the reasons
A rival could cut fares. Does your answer change?
"It could. At a 70% load the fare can only fall to about $162 before a flight stops covering its cost, so I'd want to know how likely that is before adding the second flight."
On missing data
You don't have all the numbers. Why decide?
"Because the client needs a decision. I'd say what I'd do on what we know, and name the one fact that would change it. Saying it depends leaves them where they started."
From Dartmouth Center for Career Design: case interview guidelines, Leland guide on BU Questrom's career center blog and Cornell: case interviewing.
What do interviewers listen for in your close?
BCG says there isn't always a single right answer and that what matters most is how you approach the problem and the quality of your reasoning. McKinsey says there is not one logical solution. So the close is judged on these, in our experience.
A position
One clear answer you'd stand behind. If the case is close, pick and say what would change your mind.
Reasons from this case
Cornell's guide says to explain what case facts led to your conclusion. Use the numbers you found, not general points about the industry.
Only what matters
Cornell also says to stay concise: if asked for the top two issues, give two. Leave out the analysis that didn't change the answer.
A risk and a next step
Dartmouth asks for barriers or risks. A Leland guide on BU's career blog adds one specific next step, such as a pilot test or due diligence.
From BCG: case interview preparation, McKinsey: interviewing and Cornell: case interviewing.
9 case questions to practice closing
Soreno's own questions. Work each one, then give your close in about a minute. Under each is what a strong close does.
Yes or no
The easiest close to practice: pick a side.
You've finished the analysis and the CEO walks in. What do you tell them?
Gives the answer in the first sentence and is done in about a minute. Practice it, see a model answer
Should a manufacturer build a second factory?
Says yes or no, then gives demand against capacity and payback. The risk is demand falling short. Practice it, see a model answer
Should a company make a key component itself or keep buying it?
Picks one and gives the cost per unit both ways, plus the volume at which the answer flips. Practice it, see a model answer
What's wrong and how to fix it
Close on the cause and the fix, not the analysis.
A grocery chain's profits are down 20%. What should it do?
Names the one cause that explains most of the drop, then the fix and what it's worth. Practice it, see a model answer
An airline is carrying more passengers but making less. What's your advice?
Says which of revenue or cost per passenger moved, and what to change first. Practice it, see a model answer
A streaming service's churn has doubled. What do you recommend?
Leads with who is leaving and why, then one action with its expected effect. Practice it, see a model answer
Choose between options
Say what you'd drop as well as what you'd pick.
One of your client's three divisions is losing money. Fix it, sell it or close it?
Picks one, says why the other two lost and what else breaks if the division goes. Practice it, see a model answer
A gym chain wants to double revenue in five years. How?
Ranks the options by how much of the gap each closes and says which to start with. Practice it, see a model answer
A coffee chain is thinking about entering India. What's your advice?
Gives a go or no-go, the number behind it and the first market to test. Practice it, see a model answer
Common mistakes when closing a case
The ones we hear most in practice cases. The last comes from Cornell's guide.
Retelling the case in order
The interviewer was there. Start with the decision, not with what you looked at first.
Saying it depends
Give the answer you'd act on with today's facts, then say the one fact that would change it.
Listing every finding
Two or three reasons land. Six sound like you can't tell which ones matter.
Reasons without numbers
"It's profitable" is weak. "It earns $1.4 million a year" gives the interviewer something to weigh.
A next step that could be for any case
"Do more research" says little. Name the test, who does it and what result would stop the plan.
Not checking a number that looks wrong
Cornell's guide says that if your answer seems flawed, go back and reevaluate it. Better now than after you've recommended on it.
Practice your close out loud
Soreno's AI interviewer runs a full case by voice, asks for your recommendation, pushes back on it and scores how you closed. Start your free week.
Questions people ask
How do you end a case interview?
State your recommendation first, give two or three reasons with numbers from the case, name the main risk and finish with one next step. Dartmouth's career center tells candidates to always give a concrete recommendation and the barriers or risks to it.
How long should a case recommendation be?
We aim for about a minute. A Leland coach's guide on BU Questrom's career center blog suggests practicing the summary out loud in under two minutes.
What is the difference between a synthesis and a summary?
A summary recaps what you did. A synthesis says what the client should do and why, using only the findings that changed the answer.
What if I don't have enough data to decide?
Decide anyway on what you know, and name the one fact that would change your answer. There isn't always a single right answer, BCG says, so a clear, reasoned position is what counts.
Is there a right answer to a case interview?
Usually not one. BCG says there isn't always a single right answer, and McKinsey says there is not one logical solution. Reasoning is what's judged.
Should I mention risks in my recommendation?
Yes, one or two at most. Dartmouth's guidance asks for barriers or risks, and a Leland guide on BU's career blog asks for a top risk and a specific next step.
Should I synthesize during the case too?
Yes, in our experience. After each exhibit or calculation, say in one line what it means for the question. Then the final close is mostly already said.
Can I take a moment before giving my recommendation?
In our experience, yes. Ask for 30 seconds to pull your thoughts together, write the first sentence and pick your numbers. A short pause beats a rambling start.