Product Sense
1. How would you improve Google Maps?
Start by clarifying the goal (engagement, monetization, retention?), pick a user segment, identify their unmet need, brainstorm 3-5 ideas, and prioritize by impact-vs-effort. Pick one and walk through how you would measure success.
Analytical
2. A core metric dropped 15% overnight. Walk me through how you investigate.
Segment the drop (geography, platform, user cohort, time of day) before forming a hypothesis. Distinguish data issue vs. real change. Check recent releases and external events. Communicate timeline and severity to stakeholders while the investigation is in flight.
Strategy
3. Should we build, buy, or partner for this capability?
Strategic fit, time-to-market, cost, and switching risk are the four axes. Build when it is core to differentiation; buy when speed matters more than fit; partner when you need validation before committing capital. Always name your dealbreakers.
Estimation
4. Estimate the number of Uber rides in San Francisco per day.
Bottom-up: population × Uber-using share × rides per user per day. Top-down: total trips in city × Uber market share. Compare both for sanity. The interviewer cares about the method, not the number — show your assumptions.
Execution
5. How do you prioritize when engineering capacity is constrained?
A scoring framework (RICE, ICE, or similar) plus stakeholder transparency. Name how you handle pet projects from execs and how you say no without burning the relationship. Cite a real trade-off you made.
Behavioral
6. Tell me about a product that failed and what you learned.
Pick a real failure. Be specific about what you assumed, what was wrong about it, and what changed in your process afterwards. Avoid blaming the team or the market — the interviewer is looking for ownership.
Communication
7. How do you handle a senior stakeholder pushing back on your roadmap?
Surface the underlying concern, propose a data-gathering step instead of immediately conceding or doubling down, and align on shared success metrics. Conflict resolution > conflict avoidance.