Guesstimate Questions: 10 Examples + Answers (2026)

Guesstimate interview questions in 2026 ask candidates to estimate an unknown quantity (piano tuners in a city, daily haircuts in the US, gas stations on a highway) using logic and stated assumptions rather than memorized facts. A guesstimate is a Fermi problem, and McKinsey, BCG, and Bain all use some version of it: McKinsey typically embeds a guesstimate inside a larger interviewer-led case, BCG sometimes runs it as a standalone estimation, and Bain folds market sizing into candidate-led cases. Interviewers score reasoning quality and assumption defensibility, not exact accuracy: a structured estimate that lands within an order of magnitude of the true figure typically scores higher than an unstructured guess that happens to land closer. The transferable skill is building a clean tree before touching numbers, anchoring on a reliable aggregate like US population (330M) for top-down problems, calculating without losing units, and sanity-checking the result. Confirm the exact format and time box from the interview process rather than assuming one firm-wide pattern.

What are common guesstimate questions and answers?

Below are 10 representative guesstimates spanning every major category. Each row shows the approach and answer; the three most-asked types are then worked through in detail.

These 10 examples span population, revenue, physical volume, rates, and market sizing. For market-specific variants see our market sizing step-by-step guide, or use the free case practice book to work through estimation prompts on paper. Finance candidates can add the Capital One case interview guide, where estimation turns into business analyst case math.

Sizing questions are the estimation type interviewers lean on most, so start there. Run one timed rep and get your structure and assumptions scored before you read the worked examples.

What are guesstimate questions for interviews?

A guesstimate (also called a Fermi problem, after physicist Enrico Fermi) is a structured estimation exercise where you build an unknown quantity from logic, segmentation, and reasonable assumptions. Evaluation focuses on reasoning quality and assumption defensibility, not exact accuracy (MConsultingPrep).

Every guesstimate is solvable with the same 5-step framework regardless of topic. The difference between candidates is rarely math speed; it is whether they structured the problem before naming numbers.

The 5-step guesstimate method

Every guesstimate follows the same five moves in order. Skip the structure step and you land in the most common failure mode interviewers penalize.

Execution checklist

• ✓

  Clarify scope. Lock down geography, time period, and units before doing anything else, so you estimate the quantity actually asked

• ✓

  Structure and choose your approach. Break the problem into 2-4 components and commit to top-down or bottom-up before naming a single number

• ✓

  Estimate each component. Use round anchors and state every assumption, flagging the one you are least sure about

• ✓

  Calculate aloud. Round aggressively and keep units attached so the interviewer can follow and challenge your logic

• ✓

  Sanity-check. Compare against a known reference point and adjust; a missing sanity-check is a visible red flag

For the market-sizing version of this sequence, use the market sizing step-by-step guide.

When should you use top-down vs bottom-up?

Top-down works when you have a reliable aggregate to anchor on (US population 330M, US households 130M, US cars 280M). Bottom-up works when the quantity is best built unit by unit (revenue per store × number of stores). Doing both and reconciling is the strongest signal a candidate can send (Hacking the Case Interview). For the decision rule and worked examples of each, see the top-down vs bottom-up market sizing guide.

How do you solve the three most-asked guesstimate types?

Below are full walkthroughs of a population-based, a revenue-based, and a physical-volume guesstimate. Each is the most-asked variant in its category at MBB and Tier 2 firms.

1. Piano tuners in NYC (population-based)

NYC has 8.3M people in roughly 3.3M households. Assume 5% own a piano (165,000) plus 35,000 institutional pianos (schools, theaters, churches), totaling 200,000 pianos. At 1.5 tunings per piano per year, that is 300,000 tunings annually.

Supply side: a tuner does 4 tunings per day (1.5 hours each plus travel) across 250 working days. Divide demand by capacity to get the headcount. NYC Yellow Pages historically lists 200-300, right in range.

• Anchor: 8.3M population, 3.3M households
• Key assumption: 5% piano ownership (above national due to music density in NYC)
• Sanity-check: 1 tuner per ~28,000 people, reasonable for an urban market

2. McDonald's revenue per store (revenue-based)

Build bottom-up from daily transactions by daypart. Breakfast (4 hrs × 60/hr = 240), lunch (3 hrs × 100/hr = 300), afternoon (3 hrs × 40/hr = 120), dinner (4 hrs × 80/hr = 320), late night (3 hrs × 25/hr = 75). Total: 1,055 daily transactions at $9.50 average ticket = ~$10,000/day.

McDonald's reports a US franchise average of approximately $3.7M, within 2% of our estimate.

• Anchor: 18 operating hours segmented into 5 dayparts
• Key assumption: $9.50 average ticket (varies by location and channel)
• Sanity-check: ~$10K daily revenue is reasonable for a high-traffic QSR

3. Golf balls in a school bus (physical volume)

Start from the bus interior, strip out space for seats and the engine, convert to cubic inches, then apply the packing efficiency that trips up most candidates. Published answers range 200K-500K depending on bus size, so the estimate sits squarely in the published range.

"Skipping the packing-efficiency adjustment (spheres fill only ~64% of space) is the single most common error on physical-volume guesstimates." Street of Walls 2026 case training.

• Anchor: Standard school bus dimensions (20 × 6 × 6 ft)
• Key assumption: 64% packing efficiency for randomly packed spheres
• Sanity-check: ~240K golf balls in a bus passes gut check

What are the most common guesstimate mistakes?

The five most frequent errors, per Street of Walls:

1. Calculating before structuring. Lay out your tree first.
2. Unflagged assumptions. State every number, mark the ones you are least sure about.
3. Skipping the sanity-check. A reference comparison is non-negotiable.
4. Over-segmenting. 3 segments beats 8. More segments compound error.
5. Math-test mindset. Guesstimates are reasoning exercises with arithmetic, not the other way around.

Round aggressively and state your rounding. "I am rounding 8.3M to 8M for simplicity" reads more competent than fumbling 8,300,000 x 0.05 in your head. For shortcuts see our case interview math mental shortcuts and case interview math practice, or run Quick Math to build the rounding reflex before your next mock.

Run Four Guesstimates Under a Clock

Four more prompts, one from each of the common categories. Read the prompt, work it on paper first, then check your path against the compact solution:

State the anchor, name the weakest assumption, calculate, and give one sanity check before revealing the worked answer.

Worked example: pizzas sold in Germany per day. Population: 84M. Assume the average person eats pizza twice a month, about 0.07 pizzas/day. 84M x 0.07 is about 5-6M slices-equivalent, call it 2-3M whole pizzas/day counting sharing. State the sharing assumption explicitly; frequency-based consumption guesstimates hinge on that one number, so bracket it (run once a month and four times a month) and show the range if you are unsure.

Worked example: piano tuners in Chicago, 60 seconds. Chicago metro is about 9M people, about 3.5M households, maybe 1 in 30 owns a piano, about 120K pianos. Tuned once a year is 120K tunings. A tuner does 3-4 a day across 250 days, about 900 a year. 120K divided by 900 is about 130 tuners. The scoring is on the chain of stated assumptions and clean arithmetic under time pressure, not the number itself.

Now run the same chain against a real clock and get it scored.

How do guesstimates appear inside a full case?

Standalone guesstimates are getting rarer at MBB. More often, an estimation step is embedded inside a larger case: "Before we assess market entry, size the Indian snack market" or "How many customers per day to break even?" The same 5-step framework applies, but the time budget compresses to 2-3 minutes since estimation is one node in a wider analysis.

For full worked cases that include estimation, see case interview examples, the case interview frameworks complete guide, and the market sizing questions bank.

Before your next mock, close with the version that matters most under pressure: a timed sizing rep inside the two to three minute budget a case allows.

Sources (checked June 17, 2026)

• Hacking the Case Interview guesstimate questions: hackingthecaseinterview.com/pages/consulting-guesstimate-questions
• MConsultingPrep market sizing and guesstimates: mconsultingprep.com/case-interview-market-sizing-guesstimate
• Street of Walls consulting guesstimate cases: streetofwalls.com/finance-training-courses/consulting-case-study-training/consulting-guesstimate-cases
• Wikipedia Fermi problem: en.wikipedia.org/wiki/Fermi_problem
• McKinsey careers interview prep: mckinsey.com/careers/interviewing
• IGotAnOffer market sizing: igotanoffer.com/blogs/mckinsey-case-interview-blog/market-sizing-questions