Google product sense interview questions

Sample questions with short answers

Tap a question to reveal the short answer and key points.

1. 1How would you improve Google Maps for first-time visitors to a city?HardGoogle-specificShow short answer

   Short answer

   Define the segment (tourists in an unfamiliar city), their jobs (navigate, discover, avoid mistakes) and pain points (transit confusion, language, data costs). Prioritise one — e.g. transit confidence — and propose step-by-step transit guidance with offline support. Metrics: completed trips, fewer re-routes; guardrail: no drop in search-to-direction conversion.

   Interviewers listen for

   • define the first-time visitor segment
   • prioritise one pain point
   • success and guardrail metrics
2. 2Should YouTube show more or fewer ads to a new user in their first week?MediumGoogle-specificShow short answer

   Short answer

   Frame the goal: long-term retention outweighs first-week revenue. Propose fewer or better-targeted ads early, backed by an experiment measuring week-4 retention and lifetime ad revenue, not just week-1 revenue. State the trade-off honestly and the decision rule you would set before running the test.

   Interviewers listen for

   • retention vs short-term revenue trade-off
   • run an experiment with a long-term metric
   • set a decision rule in advance
3. 3How would you improve a product you use every day?MediumAsked everywhereShow short answer

   Short answer

   Pick a product, state its goal and the user segment you will focus on, then list their pain points and prioritise one by reach and impact. Propose two or three solutions, pick one, and define a success metric plus a guardrail metric. Call out the trade-offs and how you would test it. The structure matters more than the specific idea.

   Interviewers listen for

   • define goal and user segment
   • prioritise pain points by impact
   • success metric and guardrail metric
4. 4A key metric dropped 15% week over week. How do you investigate?MediumAsked everywhereShow short answer

   Short answer

   First check data integrity — tracking changes, logging bugs, timezone or attribution shifts. Then segment: platform, geography, new vs returning users, release version. Correlate with internal events (launches, experiments, pricing) and external ones (holidays, competitors, outages). Form a hypothesis, validate it, and only then propose a fix.

   Interviewers listen for

   • rule out data or tracking issues first
   • segment the drop
   • correlate with internal and external events
5. 5How do you decide what to build next when engineering capacity is limited?HardAsked everywhereShow short answer

   Short answer

   Anchor on the product goal for the quarter, then score candidates on user value, business impact, confidence and effort (RICE or a simpler impact/effort grid). Protect a slice of capacity for tech debt and reliability. Explain the trade-offs to stakeholders with the data behind the ranking, and revisit when new evidence arrives.

   Interviewers listen for

   • tie to the quarter's goal
   • score by impact, confidence and effort
   • reserve capacity for debt and reliability