Six curveballs and how to handle them
Contradicting data, not knowing, missing data, a wrong number, a changed objective, and silence.
Key takeaways
- A curveball tests how you think when your plan breaks. Interviewers often care less that your first idea was wrong than how quickly and calmly you update it.
- Long silence on your side is the other half of this. If you need 20 seconds to compute, say so: "Give me 20 seconds to work this out." Then narrate the key step.
- The strong answer uses all four parts of the pivot and lets the data redirect the case.
- Interviewers usually forgive not knowing a fact. They rarely forgive a made-up fact stated with confidence.
- When a number is missing, turn it into a threshold: the value it would need to have for the answer to change.
This is part of the expert track. It assumes you have worked through the foundation lessons (structuring, case math, exhibits, synthesis) and several case types, and that you have run a few cases out loud. If not, start there and come back.
Key idea
A curveball tests how you think when your plan breaks. Interviewers often care less that your first idea was wrong than how quickly and calmly you update it.
The four-part pivot, said out loud
- 1What I expected: "I expected the problem to be price pressure."
- 2What the data shows: "The data shows prices are flat and store visits fell in two regions."
- 3What I now think: "So I now think the cause is local to those regions."
- 4What I will test next: "Next I would like to check whether a competitor opened there this year."
The same four parts work for most surprises: data that contradicts you, a number that is missing, a number that is wrong, or a goal that changes. Six exercises follow. Read the scenario, say your answer out loud, then compare it with the model response and the reason it works.
Exercise 1: the data contradicts your hypothesis
A German sporting goods retailer with 120 stores saw profit fall 20 percent last year. You said your hypothesis was that online rivals forced price cuts. The interviewer shows you the chart below and adds: "Average selling prices are flat, within 1 percent of last year, and online sales are flat too."
Bar chart: Change in store visits by region, this year against last year (%). North: -2 percent; South: -18 percent; East: -15 percent; West: -1 percent.
So-what
The fall is concentrated in the South and East, which points to a local cause rather than a market-wide price war.
Your price hypothesis does not match the data. What do you say?
Weaker answer
"Prices might still be the issue. Maybe there are hidden discounts we are not seeing, so I would still like to look at pricing by product."
Stronger answer
"I expected price pressure from online rivals. The data shows prices and online sales are flat, and the fall is in store visits in the South and East, down about 18 and 15 percent. So I now think the cause is local to those two regions: a new competitor, a change in access such as road works, or something in our stores there. Next I would like to know whether a competitor opened stores in the South or East this year."
Why the stronger answer wins: The strong answer uses all four parts of the pivot and lets the data redirect the case. The weak answer defends a hypothesis the data has already rejected, which wastes time and looks rigid.
Your hypothesis for a hotel chain's falling profit was rising staff costs. The interviewer says staff costs per room night are flat, but occupancy fell from 80 to 65 percent in hotels near the airport. Pivot out loud in under 30 seconds using the four parts.
Exercise 2: saying "I do not know" well
In a case about an Indian pharmacy chain moving online, the partner asks: "How do the rules on selling medicines online in India affect this?" You do not know the rules in detail.
You do not know the answer to a factual question. What do you say?
Weaker answer
"Yes, the rules say online pharmacies can only sell 20 percent of their volume online, so it is limited." (A confident claim the candidate made up.)
Stronger answer
"I do not know the current rules in detail, and I would rather not guess. Here is how I would think about it. If online sales of prescription medicines need a licensed pharmacist to check each order, then the online channel carries staff costs close to a store's, which narrows its cost advantage. I would confirm the rules with a regulatory expert in the first week. For now I will assume a pharmacist check is required, and I can show whether the answer changes if it is not."
Why the stronger answer wins: Interviewers usually forgive not knowing a fact. They rarely forgive a made-up fact stated with confidence. The strong answer admits the gap, reasons from what the rule could be, labels the assumption, and says how it would be checked.
The interviewer asks: "What margin do Brazilian supermarkets typically make?" You are not sure. Answer out loud in under 30 seconds without making up a number as if it were a fact.
Exercise 3: the data you need does not exist
A Saudi telecom operator is considering a family plan. It would cost SAR 40 million a year in marketing and systems. Each family that switches adds about SAR 800 a year of contribution, and there are about 500,000 target families. You ask what share of families would switch. The interviewer says: "We do not have that."
The one number you need is missing. What do you do?
Weaker answer
"Without that number I cannot answer. Could we assume 30 percent?" (Or silently assumes 30 percent and presents the profit as a fact.)
Stronger answer
"Then let me find the share we would need. To cover SAR 40 million at SAR 800 a family, we need 50,000 families, which is 10 percent of the 500,000. So the question becomes: do we believe more than 10 percent will switch? If the team thinks 15 to 20 percent is realistic, the plan works comfortably; if they think 5 percent, it does not. A pilot in one city could tell us within a few months."
Why the stronger answer wins: When a number is missing, turn it into a threshold: the value it would need to have for the answer to change. That makes the decision clear and shows exactly what to test.
Using the family plan numbers (SAR 40 million a year of cost, SAR 800 a year of contribution per family, 500,000 target families), what share of families must switch to break even, in percent?
Exercise 4: a number from the interviewer looks wrong
You are sizing the market for a home cleaning service in the UAE. The interviewer says: "There are 18 million households in the UAE. About 30 percent pay for cleaning, and they spend about AED 1,200 a year each." The client hopes to reach AED 500 million a year of revenue.
The household number does not look right. What do you say?
Weaker answer
Uses 18 million without comment, sizes the market at about AED 6.5 billion, and concludes that AED 500 million is an easy target. Or says "That number is wrong" in a tone that sounds like blame.
Stronger answer
"Just to check before I use it: the UAE has roughly 10 million people, so 18 million households seems too high. Did you mean 1.8 million?" Interviewer: "Yes, 1.8 million, sorry." "Thank you. Then the market is 1.8 million times 30 percent times AED 1,200, about AED 648 million a year, not AED 6.5 billion. That changes the conclusion: AED 500 million of revenue would need about 77 percent of the whole market, which is not realistic for a new entrant."
Why the stronger answer wins: A quick sense check against a known fact catches the error. The strong answer checks politely, recomputes, and, most important, says which conclusion changed. Interviewers sometimes give a wrong number on purpose to see whether you notice.
After the correction: 1.8 million households, 30 percent pay for cleaning, AED 1,200 a year each. What is the market size, in AED millions a year?
The client wants AED 500 million a year of revenue in a market of AED 648 million a year. What share of the market is that, in percent, to one decimal place?
Exercise 5: the objective changes halfway
A Spanish hotel group asked how to grow. You compared three options and recommended renovating four hotels: EUR 12 million, adding about EUR 3 million a year of cash profit once a year of works is done, the best return of the three. Then the interviewer says: "New information. The owner's bank wants EUR 15 million of debt repaid within 18 months." The other options were two new hotels (EUR 30 million, EUR 4 million a year of cash profit from year two) and selling one hotel and leasing it back (EUR 20 million of cash now, then EUR 1.4 million a year of rent).
The goal has changed from growth to cash. What do you say?
Weaker answer
"I would still renovate, because it has the best return." Or: "Let me start the case again from the beginning."
Stronger answer
"That changes the test from growth to cash within 18 months, so let me re-rank the options on that. The renovation takes EUR 12 million and returns only about EUR 1.5 million in that time. New hotels take even more cash. Only the sale and leaseback raises money: EUR 20 million now minus about EUR 2.1 million of rent over 18 months, about EUR 17.9 million, above the EUR 15 million needed. Doing the renovation at the same time would leave only about EUR 7.4 million, so I would sell and lease back one hotel now and delay the renovation until the debt is repaid. My earlier answer was right for growth; it is wrong for this new goal."
Why the stronger answer wins: When the objective changes, keep the analysis and change the test. The strong answer says the goal has changed, re-ranks the same options against the new test, and says openly that the recommendation changes and why.
A sale and leaseback brings EUR 20 million now and costs EUR 1.4 million a year in rent. What is the net cash after 18 months, in EUR millions?
If the group also renovates (EUR 12 million out, then EUR 3 million a year of cash profit for the last 6 of the 18 months), what is the combined net cash after 18 months, in EUR millions?
Exercise 6: the interviewer goes silent
You finish your recommendation. The interviewer says nothing for about ten seconds and keeps writing.
Silence after your answer. What do you do?
Weaker answer
"Actually, maybe I was wrong. It could also be the other option. Or perhaps we need more data..." (Starts taking back a sound answer to fill the silence.)
Stronger answer
Waits calmly for a few seconds. Then: "That is my recommendation. Would it help if I went further into the main risk, or into how the client would carry this out?"
Why the stronger answer wins: Silence usually means the interviewer is writing notes, not that you were wrong. Taking back a sound answer without a new fact looks like a lack of conviction. A calm pause and a short offer of what to do next keeps you in control.
Long silence on your side is the other half of this. If you need 20 seconds to compute, say so: "Give me 20 seconds to work this out." Then narrate the key step. A silent candidate is hard to score, because the interviewer cannot see how you think.
When the case surprises you, say what changed and what it means. A clean pivot often scores better than a correct first guess.
Check your understanding
Which of these is the best way to handle data that contradicts your hypothesis?
A number you need is missing and the interviewer says there is no data. What is the most useful move?
Halfway through, the client's goal changes from growth to cash. What should you do with your earlier analysis?
Sources for this lesson (1)
- Recognized public explanations of case-interview concepts and frameworks
My notes on this lesson
0 of 5,000 characters. Saves automatically.
Try the 10 remaining checks and drills above to complete this lesson (0 of 10 done).
Spotted something wrong or out of date? Report a mistake. We check every report and correct the page.