Guesstimates & Market Sizing
Part of the Product Management Roadmap
In short
"How many cups of chai are sold in Bengaluru each day?" Nobody knows the exact answer, including the interviewer. They want to see you break a big unknown into small, sensible guesses.
'How many piano tuners are in Chicago?'-style estimation questions, or full TAM/SAM/SOM prompts — testing structured, transparent quantitative reasoning under uncertainty.
Think of it like this
Guessing sweets in a jar
At a fair, you guess how many sweets are in a jar. Counting one by one is impossible. So you count one layer, count the layers, and multiply. Your answer won't be perfect, but your method will be close and easy to explain.
Watch it happen
Estimating daily chai cups sold in a city
1/5Start with a population you can defend. Assume about 1.3 crore people. State it as an assumption.
The big ideas
What you need to know
Break it down
Split the big number into a chain of smaller ones you can reasonably guess.
Say assumptions aloud
State each number and why. The interviewer grades the logic, not the exact total.
Round freely
Use easy numbers. 1.3 crore times 0.6 is fine as about 78 lakh.
Sanity check
Cross-check from another angle, like stalls times cups per stall.
See it in code
A tiny example
1=13000000 * 0.6 * 2 * 0.3The whole chain as one formula. Putting each assumption in its own cell lets you change one and see the answer update.
Save yourself the pain
Common mistakes
Oops
Freezing because you don't know the real number.
Do this instead
Say "I'll assume" and keep going. Reasonable guesses are expected.
Oops
Doing maths silently and announcing a total.
Do this instead
Walk through each step so the interviewer can follow and correct you.
Your mission
Two-way estimate
Quick check
What does an interviewer mainly judge in a guesstimate?
Go deeper
- 1Practice thinking out loud — the interviewer cares about your approach, not the exact number.
- 2Use round numbers and state your assumptions explicitly as you go.
- 3Practice 3-4 of these until the top-down/bottom-up structure feels automatic.
More analytical & metrics at this stop
Root Cause Analysis Questions
Metrics Design Questions
Stuck on this topic? Ask an Insider
Get 1:1 guidance from people who've walked this exact path — on the InsideEdge app.
