Handling Pushback in a Case Interview Without Folding

Two firms publish the instruction and neither resolves it

Read the Kearney sentence again, on its case page in a Wayback snapshot at kearney.com. It asks for two behaviors that conflict under pressure and offers no tie-breaker. That is worth noticing rather than smoothing over, because most advice on this subject silently picks a side.

L.E.K. publishes the same conflict across two of its seven numbered steps, on the preparation page at lek.com. Step three is Stick to your approach: "Outline how you are thinking and move methodically to address each issue, point-by-point". Step six is Come up with an answer: "Commit yourself to a course of action, but show flexibility and a willingness to rethink in the face of new data."

Six published instructions, two firms pulling both ways, and a third naming the failure at both ends. The subject is genuinely two-sided, which is why a page that tells you to always stand your ground or always defer is contradicting at least one firm that has written the opposite.

Handle pushback on new data, hold course on discomfort

Here is the test. It is ours, and no firm publishes anything like it.

Change course when a fact arrived that your current path cannot absorb. Hold course when nothing arrived except difficulty.

Those two situations feel identical in the room and they are not. A challenge like "are you sure the cost side matters here?" contains no new fact. A challenge like "margin rate has been flat for three years" contains one, and it kills any branch that assumed cost mix was moving.

Three tests you can run in four seconds each.

1. Name the fact. If you can say the number or the statement you just received, it is new data. If you cannot name anything, you received tone.
2. Check whether your structure still holds. A branch that is now known to be flat is dead. A branch that is merely hard is alive.
3. Ask what changes. If the answer is nothing, you are being tested on conviction. Kearney's second half applies.

L.E.K.'s step six names the trigger explicitly: rethink in the face of new data. The firm put the condition in the sentence. Every other reason to change course is discomfort wearing a fact's clothing.

One more move belongs here, and Kearney publishes it: "State your assumptions, and revisit them when the need arises." An assumption you stated out loud is one you can revise without contradicting yourself, because you already flagged it as a choice. An assumption you never voiced becomes an error when it is challenged.

How to handle pushback on a number, worked back through the chain

The most common pushback is on an input rather than a conclusion, and it has a mechanical answer. Show the chain, change one link, report the new end.

Here is a case with stated assumptions. A manufacturer runs three lines, each with a capacity of 1.2 million units a year, so capacity is 3.6 million units. Unit contribution is 6 dollars.

Recomputed by hand on 30 September 2026. Three lines at 1.2 million is 3.6 million units. At 100 percent utilization, 3.6 million times 6 dollars is 21.6 million dollars. At 80 percent, 3.6 million times 0.8 is 2.88 million units, and 2.88 million times 6 dollars is 17.28 million dollars.

Now the two pushbacks, answered out loud.

Utilization is not 100 percent. Then contribution is 17.28 million rather than 21.6 million. The gap is 4.32 million dollars, which is 20 percent of the full-capacity figure, and 20 percent is exactly the utilization shortfall. That is the check that tells you the chain is linear in utilization.

Contribution is 5 dollars, not 6. Then 2.88 million times 5 dollars is 14.4 million dollars. The fall is 2.88 million dollars, which is 2.88 divided by 17.28, or 16.67 percent. That equals one sixth exactly, which makes sense: one dollar off six is one sixth of the margin.

Notice what the second sentence does. It states the loss, states the share, and then states why the share has that value. A challenged input answered with a number and no reason invites the next challenge. A challenged input answered with a number and its mechanism ends the exchange.

The case interview signal you are told not to dismiss

Some pushback arrives as a hint rather than a challenge, and one firm publishes both the rule and a worked example.

Accenture's careers blog at accenture.com states: "If the interviewer says something, it probably means something. Don't dismiss seemingly extraneous details." Then it gives its own illustration: a retailer has been acquired, and the previous owner was fond of the children's brand. The detail signals that selling that asset is off the table.

The same page supplies the reason to treat every remark as material: "The interviewer is your biggest asset in the room. They have the information you need to "solve the case" successfully."

Bain lists the matching behavior among its nine general case tips: listen and adapt to the interviewer's signals. And BCG puts the failure on its pitfall list as Ignore Feedback, on the same page that states "there isn't always a single 'right' answer".

Put those together and a practical reading emerges. A remark that constrains the solution space is a gift. A remark that questions your reasoning is a test. Telling them apart is the skill, and the tell is whether the remark contains information about the client or about you.

How to handle pushback and still read as coachable

You can hold a position and still satisfy Kearney's first half. The trick is to make the disagreement auditable.

A three-part sentence does it. Name what you heard, name what it would change, name why you are keeping your position anyway.

Worked, on the utilization challenge above: if utilization were 60 percent rather than 80, contribution drops to 12.96 million and the investment case fails. I am staying with 80 because the plant ran at 82 percent last year, and I would want the maintenance schedule to confirm it. Verified: 3.6 million times 0.6 is 2.16 million units, and 2.16 million times 6 dollars is 12.96 million dollars.

That sentence does three things at once. It takes the challenge seriously enough to compute it. It shows the consequence if the challenger is right. And it gives a reason and a test rather than a preference.

The version that fails is shorter and more common. I still think 80 percent is right. No computation, no consequence, no test. That reads as the thing BCG calls Ignore Feedback, even when the position is correct.

The other failure is the opposite. Accepting the challenger's number without comment abandons the chain you built, and Deloitte's own framing of the exercise as a business conversation is worth holding onto here: a colleague who changes a number silently is harder to work with than one who argues.

What holding and folding cost in our weights

The numbers below belong to us. They are read from the grading code in src/lib/grading, and no consulting firm publishes a marking scheme of any kind.

Rehearsing the challenge

Nobody gets better at this by running cases that go smoothly. Build the challenge in.

1. Have your partner challenge one input per case, chosen at random. Your job is the three-part sentence: heard, consequence, decision.
2. Have them challenge one input where they are right. Practice yielding cleanly, which means naming the fact and restating the conclusion it forces.
3. Have them make one remark that is a hint rather than a challenge. Accenture's child-brand example is the template. Score yourself on whether you used it.
4. Say every assumption out loud before you use it. Kearney publishes this as an instruction, and it is what makes a later revision look like method.
5. Time your response to a challenge. Under four seconds reads as reflex. Over twenty reads as the pitfall BCG calls Panic Under Pressure.

Our math drill reports Working separately from Answer, which lets you see what a visible chain is worth when an input moves, and the structure drill will score a revised structure fresh. Both are free and neither needs an account. Neither can push back at you, which is the part that needs a person on the other side of the table.

For the wider composure problem, see case interview pressure. For what to do when you have lost the thread entirely, see recovering mid-case. For how the same firms describe speaking during a case, see communicating in a case.