Product sense framework

Product sense interviews test whether you can find a user, name a real pain, propose a solution that fits a company's goals and defend trade-offs under time pressure. Meta calls it product sense; Google often says product design; startups may just say "how would you improve X?" Interviewers are not looking for the "right" app idea. They are looking for whether they would trust you in a messy product discussion tomorrow.

Do not memorize a script. Internalize a sequence and waypoint — say what you will do next so the interviewer can redirect. "I'll segment next — stop me if you want a different cut" invites collaboration and saves you from a 15-minute wrong path.

The big idea

Most loops mix design + a metrics follow-up. End design answers in how you would measure success. If time runs out, skip a third idea before you skip the metric.

1. Clarify (2–3 min). Geography, platform, company context, timeline, success definition. State assumptions when they stay vague: "I'll assume we're Meta, US-first, mobile-first, success is weekly active volunteers matching to opportunities."
2. Goal and north star (2–3 min). Tie to mission, then pick a north star. "Confirmed volunteer hours per weekly active user — not app downloads."
3. Users (5–8 min). List 3–4 meaningfully different segments. Score size, underserved-ness, fit to company strengths. Pick one. Weak: "Users who like volunteering." Strong: "25–40 urban professionals who already donate money but fail to convert intent-to-volunteer because discovery of nearby 2-hour shifts is broken."
4. Pains (5–8 min). Walk discover → decide → attend → reflect. List 3, prioritize frequency × severity × ability to win. Pick one.
5. Solutions (8–10 min). Three distinct approaches — not three flavors of a button. Dimensions: social graph vs marketplace vs content; AI assist vs human curation; supply vs demand. Choose one.
6. MVP, metrics, trade-offs (5 min). V1 scope, explicit non-goals, input metrics, guardrails, biggest risk and how you'd learn in 4–6 weeks.

Interviewers reward decisions. Listing ten ideas without picking is a fail. Picking with a crisp "because" against your north star is the skill.

Walking through it

Interviewer: "Design a travel product for Meta." Weak answer invents a generic TripAdvisor clone. Strong answer picks group trip planning among existing friends (Messenger/WhatsApp graph), pain = coordinating dates and budgets in 40-message threads, solution = shared trip object with polls + payments split + destination shortlist, metric = trips with ≥3 contributing members per week, guardrail = spammy invites. You used Meta's unique asset and made a cut.

Common mistakes

• Boiling the ocean. "There are many users…" for 10 minutes. Pick a segment by minute 8.
• Feature salad. Ten incremental UI ideas, no strategy. Force three different solution shapes.
• No decision. "We could A/B all of them." Choose; offer what you'd test second.

Check yourself

1. Time is almost gone and you have not named a metric. What do you cut?

A third brainstorm idea. Interviewers remember an empty ending more than a missing option.

2. Why pick one segment instead of covering everyone?

Depth on one segment beats shallow coverage. Interviewers can always ask you to expand.

Key takeaways

• Run clarify → goal → segment → pain → solutions → MVP → metrics; waypoint as you go.
• Three distinct solutions, then a choice you can defend against the north star.
• End design answers in metrics and trade-offs.

Further learning

Lenny's definitive guide to product sense interviewsA step-by-step playbook to help you ace your PM interviewlennysnewsletter.com

Next on the map: product sense interviews.