When the interviewer asks for your recommendation, answer the client’s decision. Choose the action the evidence supports, explain the findings that make it sensible and identify the next step that matters.
I would rather hear one decisive reason developed properly than three reasons added to fill a template. A failed investment hurdle may settle a no-go. A market-entry decision may require demand, economics and delivery evidence together. Let the decision determine the support.
Your closing answer should make it easy for the client to understand what they are being asked to approve.
Return to the original decision
Before speaking, check the objective and constraints at the top of your notes. Was the client trying to increase revenue, contribution, service coverage or cash return? Was the decision about a pilot, a full rollout or a particular segment?
A recommendation to grow sales does not answer a profit question unless the economics follow. A successful urban pilot does not automatically support every rural location.
If the case has changed the scope, make that visible in the first sentence:
“I would roll out the change to the tested urban routes and keep the rural routes outside this release.”
That is a clearer commitment than “I would proceed, subject to some risks.”
Build the answer around four useful questions
Use these questions while preparing your close:
| Question | What the listener needs |
|---|---|
| What should we do? | The action, scope and timing of the commitment |
| Why? | The findings that make this choice preferable or necessary |
| What could change the answer? | The material condition or unresolved assumption |
| What happens next? | A specific action that implements or tests the decision |
You do not need to announce these as four headings. They are a way to check that your answer is complete.
Use the findings to recommend an action. The client needs to know what the analysis means for the decision.
Work through a complete no-go recommendation
Suppose a fictional charging expansion requires £450,000 of investment and generates £90,000 of annual net cash. The client requires simple payback within four years.
Payback is £450,000 ÷ £90,000 = five years. At the same investment, meeting four years requires £112,500 of annual net cash. At the same cash generation, investment must be no more than £360,000.
A complete answer could be:
“I would not approve the charging expansion on the current base case. It pays back in five years, outside the four-year requirement. I would reopen the decision if validated demand can raise annual net cash to at least £112,500, or if investment falls to £360,000 at the current cash level. The next step is to test whether contracted fleet demand or a lower installation quote can achieve either threshold.”
The answer gives a decision now and a concrete route to reconsideration. It does not turn a required improvement into an assumed forecast.
Rehearse the recommendation once with the evidence available, then again after a lower-demand scenario. Explain whether the decision changes.
Quick Answer Repair Drill
A marina operator asks whether to build 40 new berths. A candidate says no on payback, then ends: "I would revisit it if things improve."
- Investment
- £600,000
- Annual net cash
- £100,000
- Board payback limit
- Within five years
Which reopening condition should replace the candidate's final sentence?
Select the findings that actually earn the choice
Imagine you also learned that the charging market is growing, customers like the brand and the site has good road access. Those observations may matter to the wider opportunity, but they do not override a binding four-year hurdle in the supplied case.
If you include them, explain their role: they justify investigating an improved version, not approving the version that fails.
Likewise, if two options both meet the hurdle, explain why one is preferable. It may create more net value, serve the required customer group or preserve flexibility. “Both are attractive” leaves the choice unfinished.
Check that your closing sentence answers the original objective. You do not need to repeat every branch you explored to make that connection.
Make the remaining risk change an action
“Demand may be lower” is incomplete. Say what the client should do because demand matters.
“Before committing to all sites, I would validate the fleet contracts that produce the required annual cash. If that threshold cannot be reached, retain the current network.”
A next step should resolve the uncertainty, limit the commitment or make implementation work. “Do more research” and “monitor closely” leave the client to design the important part.
You also do not need to list every possible concern. Choose the one that could reverse the choice or prevent the plan from working. Name the evidence that would resolve the risk and the action you would take if the result crossed your decision threshold.
Quick Recommendation Drill
A regional airline plans a new weekday route for energy-sector business travellers. In the base case, the route clears the board's return target.
- Return target needs
- 60% of seats filled
- Client firms' travel forecasts
- 45% to 75% of seats filled
- Aircraft lease
- 12-month minimum, committed at signing
Which next step should lead the candidate's final recommendation?
Keep a recommendation ready before the case ends
After each major calculation or exhibit, add a short implication to your notes: “fails payback,” “urban segment works,” “demand exceeds usable capacity.” Those notes make the final answer easier to assemble than a page of unlabelled numbers.
If the interviewer asks you to conclude before every branch is complete, state the decision supported so far and name the unresolved part. For example:
“I would reject the current investment on its financial hurdle. We have not finished comparing alternative sites, so I would not yet recommend which replacement location to choose.”
You have still answered what can be decided. You have not invented evidence to make the ending feel complete.
Practise the close as a separate skill
Take a completed case and give the recommendation in one minute as a rehearsal exercise. Ask a listener to repeat the action, the strongest support and the condition that changes it. Then remove a piece of evidence or change a number and give the answer again.
The second attempt tests whether you can reason your way to a different recommendation rather than deliver a memorised ending.
Try Manufacturing Downtime. State the decision first, support it with the relevant case facts and leave the client with a next action they can understand.
Final recommendation drill
Would your recommendation convince an MBB interviewer?
Manufacturing Downtime
Move from the charging illustration to a manufacturing decision. Explain your recommendation to Casey, Case Room’s AI interviewer, then review feedback on evidence and next steps.
Frequently asked questions
Framing the recommendation
What should I check before I start my final recommendation?
Check the client's objective and constraints at the top of your notes before you speak. Confirm whether the goal was revenue, contribution, service coverage or cash return. Also confirm whether the decision covered a pilot, a full rollout or one segment. A recommendation to grow sales does not answer a profit question unless the economics follow.
Can a successful pilot justify recommending a full rollout?
Not automatically. A pilot shows what worked under its own conditions, so check whether the decision concerns that segment or every location. In the article's example, a successful urban pilot does not automatically support every rural location. Limit the recommendation to where the evidence applies, or name what must be validated before you extend it.
What structure should my closing answer follow?
Work through four questions: what should we do, why, what could change the answer, and what happens next. The first covers the action, scope and timing of the commitment. The second gives the findings that make the choice preferable or necessary. The third names the material condition or unresolved assumption. The fourth points to the next step that matters.
How many reasons should I give to support my recommendation?
Give as many as the decision needs, and sometimes one is enough. One decisive reason developed properly is stronger than three reasons added to fill a template. A failed investment hurdle may settle a no-go on its own. A market-entry decision may need demand, economics and delivery evidence together.
Supporting the choice with evidence
How do I present a no-go recommendation using the numbers?
Show the hurdle the option fails and what it would take to pass. In the article's fictional charging case, £450,000 of investment and £90,000 of annual net cash give a five-year payback against a four-year requirement. Meeting four years needs £112,500 of annual cash at the same investment. Alternatively, investment must fall to £360,000 at the same cash.
Why calculate break-even figures if the answer is already no?
They turn a rejection into something the client can act on. Showing that four-year payback needs either more annual cash or lower investment tells the client what an improved version must achieve. The figures also show the interviewer how far the option falls short of the hurdle, not just that it fails.
What should I do with positive findings that don't support my recommendation?
Include them only if you explain their role. In the charging example, market growth, brand appeal and good road access do not override a binding four-year payback hurdle. They can justify investigating an improved version of the investment. They should not be presented as reasons to approve the version that fails.
Risks, next steps and timing
Is it enough to say ‘demand may be lower’ as my main risk?
No, because it names a worry without telling the client what to do. Link the risk to an action. For example, validate the fleet contracts that produce the required annual cash before committing to all sites. If that threshold cannot be reached, retain the current network. The risk then has a test and a consequence.
What counts as a good next step in a case close?
A good next step resolves the uncertainty, limits the commitment or makes implementation work. ‘Do more research’ and ‘monitor closely’ fall short because they leave the decision where it was. Validating specific contracts against a cash threshold is better. It tells the client exactly what to check and what follows from each result.
What if the interviewer asks me to conclude before I've finished the case?
State the decision the evidence supports so far and name the part that is still unresolved. Give a provisional answer with a clear limit rather than refusing to conclude. This is much easier if you have already noted the implication of each calculation as you went. You can then assemble the answer quickly.
How should I take notes during the case to make the close easier?
After each major calculation or exhibit, write a short implication beside it. Examples from the article include ‘fails payback’, ‘urban segment works’ and ‘demand exceeds usable capacity’. These labels record what each result means for the decision. Your final recommendation then becomes a matter of assembling conclusions, not reinterpreting a page of unlabelled numbers under pressure.
Practising the close
How can I practise giving case recommendations?
Treat the close as its own rehearsal exercise. Take a completed case and give the recommendation in one minute. Then ask a listener to repeat three things: the action, the strongest support and the condition that would change it. If they cannot recall one of those, that part of your close needs to be clearer.
How do I stop my recommendation sounding memorised?
Change the inputs and answer again. Remove a piece of evidence or alter a number, then give a fresh recommendation. This tests whether you can reason towards a different conclusion when the facts change, rather than repeat a script. If your answer stays the same after a decisive number moves, revisit how you built it.