Skip to content

Market Sizing Framework for Case Interviews

Market sizing questions ask you to estimate something large and specific — "how many pianos are sold in the US each year?" — with no real data in front of you. The skill isn't guessing well; it's breaking the number into pieces small enough to estimate honestly.

Two approaches

Top-down: Start from a large, roughly-known number and narrow it with a chain of assumptions. "How many coffee shops are in this city?" → city population → % who drink coffee out → visits per week → coffee shops needed to serve that demand.

Bottom-up: Build the estimate up from one small, concrete unit. "How many coffee shops are in this city?" → average revenue per coffee shop → total coffee spend in the city → divide.

Neither is universally better — pick whichever gives you numbers you can actually estimate with some confidence, rather than forcing one approach because it's the one you remember.

The structure, step by step

  1. Clarify exactly what's being asked — annual sales? Current installed base? A specific region? Getting this wrong invalidates the whole estimate.
  2. Pick top-down or bottom-up based on which gives you more reliable starting numbers.
  3. Break the number into a chain of estimable factors — each one MECE with the others, so you're not double-counting or missing a step.
  4. State each assumption out loud as you use it — "I'll assume roughly 60% of adults drink coffee regularly" — so the interviewer can follow your logic and correct you if a number is far off.
  5. Do the math cleanly, rounding to numbers you can actually multiply in your head.
  6. Sanity-check the final number — does it feel roughly right compared to something you already know?

A worked example

How many piano tuners are there in Chicago?

  • Chicago population: ~2.7M → ~1M households
  • % of households owning a piano: ~2% → ~20,000 pianos
  • Tunings needed per piano per year: ~1 → 20,000 tunings/year
  • Tunings one tuner can do per year: ~500 (2/day, 250 working days)
  • 20,000 ÷ 500 ≈ 40 piano tuners in Chicago

Every step is a small, defensible assumption — the estimate is only as good as the chain, and the chain is only trustworthy if each link is MECE and stated clearly.

Practice building your own chains

Market sizing is one of the fastest frameworks to get comfortable with through repetition — the structure (not the specific numbers) is what transfers between questions. Practice case structuring on Meceify →

Turn this into a tree

Take what you just read and build it out — no sign-up required, saved automatically in your browser.

Start building for free

FAQ

What's the difference between top-down and bottom-up market sizing?

Top-down starts from a large known number (total population, total market) and narrows it down through a chain of percentages. Bottom-up builds the estimate up from a small, concrete unit, like revenue per store or per customer, multiplied out.

Do market sizing answers need to be exactly right?

No. Interviewers care about the structure of your estimate and whether your assumptions are reasonable and clearly stated, not whether you land on the precise real-world number.