SkilsMVP
BootcampInterview prep

Interview Drills — Analytical & Estimation

6 drills with frameworks and rubrics.

Interview Drills — Analytical & Estimation

"Guesstimate" / market-sizing and analytical questions. They test structured reasoning under uncertainty, not the exact number. Each has a Framework, a Model answer, and a Rubric. State assumptions out loud, use round numbers, and sanity-check the result.

The universal estimation structure: Clarify the question → break it into a few drivers → estimate each with stated assumptions → combine → sanity-check the magnitude. Interviewers care about the approach and your assumptions, not precision.

D1

  • difficulty: medium
  • concept: product-sense-interview How many cups of coffee are sold in your city per day?
  • Framework: Start from population → fraction who drink coffee daily → average cups per drinker → add a rough adjustment for visitors/multiple purchases → sanity-check.
  • Model answer: "Say a city of 1M. ~50% drink coffee daily = 500k. Average ~1.5 cups bought out per day among those = ~750k. Round to ~700k–800k cups/day. Sanity check: that's well under one per resident, which feels right."
  • Rubric: Strong answers state population and each assumption clearly, multiply cleanly, and sanity-check the magnitude. Weak answers blurt a number with no structure, or get lost in false precision.

D2

  • difficulty: medium
  • concept: product-sense-interview Estimate the number of gas stations in the United States.
  • Framework: Population → cars → cars served per station → derive count; or top-down from a known anchor. State assumptions; sanity-check.
  • Model answer: "~330M people, ~1 car per 1.5 people ≈ ~220M cars. Assume one station serves ~2,000 cars. 220M / 2,000 ≈ 110,000 stations. Sanity check: real figure is ~110–150k, so the approach lands in the right range."
  • Rubric: Strong answers pick a sensible driver, state the cars-per-station assumption, and reach a defensible range. Weak answers guess with no derivation or use wildly off anchors without noticing.

D3

  • difficulty: medium
  • concept: product-sense-interview How would you size the market for a new meal-kit subscription?
  • Framework: Define the target customer → estimate how many exist (TAM/SAM via population funnel) → realistic adoption % → price → annual revenue. Distinguish total market from realistically reachable.
  • Model answer: "Target: busy urban households that cook. From population → urban → households that cook regularly → can afford a premium kit, narrow to, say, a few million households. Assume a low single-digit % adoption × annual spend → a revenue range. I'd separate the total addressable market from what we can realistically capture early."
  • Rubric: Strong answers build a funnel from population to a reachable segment, apply a realistic adoption rate and price, and separate TAM from reachable. Weak answers conflate "everyone" with the market or skip adoption/price.

D4

  • difficulty: easy
  • concept: core-metrics-vocab What metrics would you track for a ride-sharing app?
  • Framework: Anchor on a North Star (real value) → cover the funnel/marketplace both sides (riders and drivers) → name a counter-metric → avoid vanity metrics.
  • Model answer: "North Star: completed rides per week. Riders: activation (first ride), retention, wait time. Drivers: active drivers, utilization, earnings. Counter-metric: cancellations / complaints. I'd avoid vanity metrics like total app downloads."
  • Rubric: Strong answers pick a value-based North Star, cover both sides of the marketplace, and include a counter-metric. Weak answers list vanity metrics or ignore one side of the market.

D5

  • difficulty: hard
  • concept: core-metrics-vocab Estimate the annual revenue of a streaming service with 50M subscribers.
  • Framework: Subscribers × blended monthly price × 12, adjusting for plan mix / churn / regional pricing; state assumptions; sanity-check.
  • Model answer: "50M subs × ~$12 blended monthly = $600M/month × 12 ≈ $7.2B/year. Adjust down slightly for cheaper regional/ad tiers, so ~$6–7B. Sanity check: per-sub annual ~$130–150, which is reasonable for streaming."
  • Rubric: Strong answers use a blended price, annualize correctly, adjust for plan mix, and sanity-check per-sub value. Weak answers use a single naive price, make arithmetic errors, or skip the sanity check.

D6

  • difficulty: hard
  • concept: product-sense-interview How many photos are uploaded to a large social app per day?
  • Framework: Active users → fraction who post daily → average uploads per poster → combine; sanity-check against a known scale.
  • Model answer: "Say 500M daily active users. ~20% upload a photo on a given day = 100M posters × ~2 photos = ~200M uploads/day. Sanity check: that's a believable order of magnitude for a large platform; I'd refine the posting rate with real data."
  • Rubric: Strong answers separate active users from posters, apply a posting rate and uploads-per-poster, and sanity-check the order of magnitude. Weak answers multiply unrealistic rates or never check plausibility.