Product sense interview questions for product managers

"Design a product for…" and "How would you improve…" questions test whether you can pick the right user, find the problem that matters and make a clear call. Here's how the round is graded, a method that works, a worked example and 17 cases you can practise as free mock interviews.

What the round tests

Product sense (also called product design) rounds give you an open brief and very little data. The panel isn't looking for the most features. It's looking at your judgement: why this user and not another, which problem is worth solving first, which idea you'd actually build, and how you'd know it worked. Strong candidates narrow early and explain every choice. Weak ones list ideas for every user at once and never commit.

How product sense is graded

No company publishes its rubric, but interviewer and candidate accounts agree on what panels look for. Most grade each area on its own, so a brilliant idea doesn't make up for weak user choice.

Clarify and plan: two or three questions that change the answer, stated assumptions, and a plan you check in on.
Why this company: one line on why this company should build it, used to guide later choices.
Users and segments: the players involved, split by behaviour or motivation rather than just age or income.
Choosing a segment: one segment with stated reasons (size, how underserved, fit with the goal), not simply the biggest group.
Pain points, not wishes: concrete frictions from the user's journey, ranked by how often and how badly they hurt.
Range of ideas: three or more genuinely different solutions, including one bold but buildable idea.
Prioritisation and a first version: a clear pick using impact and effort, described well enough that a team could start.
Metrics: a north-star metric tied to the user problem, a few input metrics and a guardrail that catches harm.
Trade-offs and risks: what the design gives up and for whom, plus privacy, abuse, reliability or cost risks and how to reduce them.
Vision and scale: where it could be in a few years, and how it works for a large, varied user base.
Composure: taking pushback and new facts calmly, and changing the answer when the evidence says so.
AI fluency (AI rounds): what models can and can't do, designing for wrong answers, and the cost and speed of each call.

Synthesised from public prep guides including Lenny's Newsletter, Product Alliance and Prepfully.

The method, step by step

Agree the goal and constraints: why the company wants this and what success means.
Choose one user segment and say why it matters most.
List their pain points and rank them.
Generate at least three different solutions.
Prioritise one by impact and effort, and describe a first version.
Define a north-star metric, a guardrail, and what you'd build later.

A worked example answer

The question: "Design a product that helps people who've just moved to a new city build a social circle." (Practice scenario, not one of our cases.)

Goal. Assume a new standalone app launching in one metro with a small team. Success means newcomers meet people in person regularly, not just chat.
Users. Students moving for college, young professionals moving for a first job, families relocating, and older people moving closer to children. I'd pick young professionals aged 22–30: the largest group moving into big tech cities, few existing contacts, evenings and weekends free, and loneliness peaks in their first months.
Pain points, ranked. (1) Not knowing where to start, and it feels awkward to approach strangers. (2) Plans fall through and nobody follows up. (3) Safety worries about meeting people they don't know. (4) Colleagues don't share their interests.
Ideas. (a) Small-group weekly meetups of 4–6 people matched by interest and neighbourhood, with the venue booked for them. (b) "Plus-one" invites to existing friend groups' plans. (c) Intros through employers' onboarding for new joiners. (d) Skill-swap classes run by members.
Prioritise. Small-group meetups solve the top two pains at once: the app starts the conversation and fixes the time and place. Public venues and verified profiles cover safety. Plus-one invites depend on hosts we don't have yet, and employer intros need sales, so they come later.
First version. One city, three interest tracks (food, sport, games), one weekday dinner a week, groups formed automatically, reminders on WhatsApp, a two-tap "meet again?" after each dinner.
Metrics. North star: share of new members who attend two or more meetups in their first four weeks. Guardrails: safety reports per 1,000 meetups and no-show rate.
Trade-off. Hand-picked venues and careful matching limit how fast we can grow. I'd accept slower growth early, because repeat attendance is what makes friendships, and it's what would make people recommend the app.