Your interviewer asks:
“How many takeaway coffees are sold near your university each week?”
You reach for your calculator.
That is usually too early.
The first mistake happens before the first number.
Market sizing looks like a math exercise. In reality, the arithmetic is often the easy part.
The difficult part is deciding what deserves to be multiplied.
Your first question should shrink the ambiguity
Before calculating anything, ask what exactly you are estimating.
One campus or the surrounding neighborhood?
Students only or everyone?
Coffee sold at cafés only, or convenience stores too?
One academic week or an average week across the year?
These questions are not stalling.
They define the market.
A precise calculation built on the wrong definition is still wrong.
Clarity comes before arithmetic.
Once the scope is clear, your next job is to choose a route. You could estimate from the population down, or estimate from shops and transactions up.
Both can work.
The question is which route gives you assumptions you can defend.
For the core approaches and structures, see Market Sizing.
The first 10 questions are really a sequence
You do not need to fire ten questions at the interviewer.
Think of them as ten checks your reasoning should survive:
- What exactly am I sizing?
- What geography is included?
- What time period matters?
- Who is the relevant customer?
- What is my simplest equation?
- Which assumption matters most?
- Can I estimate that assumption credibly?
- Does another route produce a similar answer?
- Is the final number plausible?
- What does the estimate imply?
Notice what question ten does.
It turns a calculation into a case answer.
Suppose you estimate 50,000 coffees per week.
Do not stop there.
What could the client do with that information? Is the market large enough to enter? How much share would a new café need? Does the estimate support another location?
A number without implication is unfinished.
Your assumptions are where the interview really happens
Students often worry that they do not know the exact population, purchase frequency, or average spend.
You usually will not.
That is the point.
The interviewer is watching how you deal with uncertainty.
| Weak move | Better move | Why it matters |
|---|---|---|
| Guess “20,000 students” | State why 20,000 is plausible | Makes reasoning inspectable |
| Assume everyone buys coffee | Segment coffee drinkers from non-drinkers | Avoids hidden overestimation |
| Use 7 purchases per week | Separate weekday and weekend behavior | Connects assumption to reality |
| Accept the final number | Sense-check from another angle | Catches structural errors |
There is another subtle distinction: an assumption is not the same thing as an estimate.
If you blur the two, your logic can look much weaker than your arithmetic. Assumptions vs Estimates in Case Interviews: Why Mixing Them Up Hurts Your Logic goes deeper on that difference.
The calculator should arrive later than you expect
Once the structure is sound, the math becomes mechanical.
Population × relevant share × frequency.
Stores × transactions per hour × operating hours.
Households × penetration × annual purchases.
Nothing magical.
The sophistication comes from choosing the right equation, defending the assumptions, and recognizing when the answer looks absurd.
So on your next market sizing case, resist the urge to impress the interviewer with speed.
Ask the question that changes the equation first.
Then another.
Then calculate.
Your first ten questions build the answer.
The multiplication only reveals what you built.