Case Study / Guesstimate Questions
Part of the Data Analyst Roadmap
In short
"How many cups of chai are sold near your college every day?" There's no right number. The interviewer wants to see if you can break a fuzzy problem into sensible pieces.
'How many X are there in Y?' or 'walk me through analyzing this business problem' — testing structured thinking under ambiguity, similar in spirit to product management case interviews.

Think of it like this
Counting guests at a wedding tent
Your aunt asks how many people came to the wedding. You don't count heads. You count about 25 rows of chairs, guess 20 chairs per row, and add a few standing near the food. A guesstimate is that quick, structured counting.
Watch it happen
Chai cups near a college gate, per day
1/5Restate the question, and ask: one college or all colleges? Weekday or weekend? Say your choice.
The big ideas
What you need to know
Clarify scope
Ask about place, time and what counts. Then state your choices out loud before calculating.
Break it into factors
Use a tree: population × share × frequency. Each factor should be something you can reason about.
Round numbers freely
Use easy numbers like 5,000 and 40%. Interviewers care about logic, not decimals.
Sanity check the end
Does 3,500 cups feel right for a few stalls? If each stall sells 500 a day, that's about seven stalls. Plausible.
Save yourself the pain
Common mistakes
Oops
Throwing out one big number with no breakdown.
Do this instead
Show the tree, state each assumption and multiply step by step.
Oops
Getting stuck trying to find the exact true value.
Do this instead
Say "I'll assume X" and move on. Mention how you'd check it with real data later.
Your mission
Three guesstimates in a row
Quick check
What is the interviewer mainly judging in a guesstimate question?
Go deeper
- 1Practice a repeatable structure: clarify the goal, break the problem down, estimate, sanity-check.
- 2Think out loud and use round numbers rather than false precision.
- 3Practice 3-4 of these until the structure feels automatic, not improvised each time.
More statistics & analytical at this stop
Stuck on this topic? Ask an Insider
Get 1:1 guidance from people who've walked this exact path — on the InsideEdge app.