What is the market sizing wheel?

A free timed challenge for case interview practice. You spin for a random market sizing question, get 3 minutes to structure it and 3 minutes to present it, then type your number to see how close you came to a solid range.

How close does my number need to be?

Inside the range is a strong answer. Within double or half of the range still counts as close, because interviewers grade the logic more than the number. Past that, one of your assumptions is off by a wide margin and the worked answer will show which one.

Does the wheel grade my structure?

No. The wheel checks the number only. To get your structure, assumptions and sanity check scored, use the AI-graded market sizing drill linked from the result.

Can I change the 3-minute clocks?

You can pause either clock or skip ahead, but the length is fixed at 3 minutes each so every round is comparable.

Market Sizing Wheel

EasyMediumHardMixed

Spin
Structure 3:00
Present 3:00
Check

Spin for a random question.

No AI. No calculator. Pen and paper.

The quick version: The market sizing wheel is a timed challenge: one random question, 3 minutes to structure it on paper, 3 minutes to present it out loud, then one number to commit to. The check tells you how many times too high or too low you landed.

Why practice market sizing with a random question and a clock?

Practising alone makes market sizing too easy in two ways: you pick a question you already like, and you take as long as you want. An interviewer gives you neither. The wheel takes both away.

A random draw trains the part that matters most, which is choosing an approach for a question you did not expect. The two clocks train the rest: getting a structure on paper before the math, then talking through it without drifting.

The number check at the end is deliberately blunt. It compares your one number with a solid range and tells you the multiple you missed by. It does not judge your structure. For that, run the same kind of question in the AI-graded market sizing drill, which scores structure, assumptions, math and the sanity check.

How the market sizing challenge works

Pick a difficulty and spin. Easy questions are a short chain from a population. Hard ones take more steps and more judgment calls.
Structure for 3 minutes. Pen and paper only. Write the steps and the assumption behind each one before you compute anything.
Present for 3 minutes. Out loud, as if the interviewer were in the room. State each assumption, do the math in round numbers, land on one number.
Check your number. Type it in. You see where it sits against the solid range, then the worked answer.

The rules that make it a fair test: no AI, no calculator, no looking anything up.

View worked examples →

3 worked

3 questions from the wheel, worked

One easy, one medium, one hard. Try each against the clock before you open the answer.

Example 1
How many cups of coffee are sold daily in Manhattan?
Approach
Population-based: Start with Manhattan population, estimate coffee drinkers, cups per person per day.
Answer
Manhattan daytime population: ~3M (includes commuters)

Coffee drinkers: 3M × 60% = 1.8M people

Cups per person per day: 1.5 (many drink 2+)

Total cups: 1.8M × 1.5 = 2.7M cups/day
Example 2
How many gas stations are there in the United States?
Approach
Demand-based: Estimate cars, gas consumption, and station capacity.
Answer
Total cars: ~280M

Annual miles per car: 12,000

Gallons per car per year: 12,000 / 28 MPG = 430 gallons

Total gallons per year: 280M × 430 = 120B gallons

Gallons per day: 120B / 365 = 330M gallons/day

Stations needed: 330M / 4,000 gallons per station = ~82,000 stations
Example 3
How many piano tuners are there in Chicago?
Approach
Demand-based: Estimate pianos, tuning frequency, and tuner capacity.
Answer
Households: 3.5M

Pianos: 3.5M × 5% = 175,000 pianos

Tunings needed per year: 175,000 × 1.5 = 262,500 tunings

Tunings per tuner per year: 4 × 220 = 880

Tuners needed: 262,500 / 880 = ~300 piano tuners

View common mistakes →

3 pitfalls

3 mistakes the clock exposes

Starting the math in the first minute
Use the structure clock for structure. If your page has numbers on it before it has steps, you will spend the present clock repairing the logic.
Presenting a range instead of a number
The check asks for one number because an interviewer does too. Say the number, then say which assumption you trust least.
Spinning again when the question looks hard
The question you want to skip is the one worth doing. Take it, miss by 5x, read the worked answer, and you will not miss that pattern again.