Product manager interview questions — and the CV claims behind them
The product, behavioural and metrics questions PMs actually get, what each is really probing, and which line on your CV an interviewer will press hardest.
What this interview is actually testing
Whether you make decisions or describe them — and whether the outcomes on your CV are outcomes you caused or outcomes you were nearby for.
The questions, and what each one is really asking
1.Tell me about a product decision you got wrong.
What it probes: Whether you can locate the decision. Weak answers blame a dependency, a reorg or engineering capacity. Strong ones name the judgement call, what information was available at the time, and what signal was ignored.
2.How did you decide what to build next?
What it probes: Whether prioritisation was a framework you can name or a process you actually ran. Interviewers listen for what got cut and who was unhappy about it.
3.What metric did you own, and what moved it?
What it probes: The line between influence and attribution. Almost every PM CV claims a metric improved; far fewer PMs can explain what else was happening in the business at the same time.
4.Walk me through how you'd improve a product I use.
What it probes: Whether you ask who the user is before proposing features. The candidates who start listing ideas have told you how they run discovery.
5.Tell me about a time you disagreed with engineering.
What it probes: Whether you can represent the engineering argument accurately. A PM who describes every technical objection as resistance is describing a team that stopped telling them things.
6.How do you know a feature succeeded?
What it probes: Whether success was defined before launch or narrated after it. The strongest answer includes a case where the definition was met and the feature was still a mistake.
7.What did you kill?
What it probes: Seniority, faster than any other question. Shipping is table stakes; stopping something with momentum behind it is the part that requires standing.
The CV claim they'll press hardest
An outcome metric — "grew activation 30%", "increased retention by 12 points".
Product outcomes are almost never single-cause, and interviewers who have run product know it. The follow-up is not "is that true?" but "what else was happening?" — a pricing change, a marketing push, a seasonal effect, three other teams shipping. A PM who volunteers the confounders is more credible, not less, because it shows they read their own metrics rather than harvesting them. A PM who claims sole credit for a number the whole company moved has just told the interviewer how they will report to them.
Product management interviews have a specific failure mode, and it is not being underprepared. It is being fluent.
PMs are professionally good at narrative. The job is partly to make a messy quarter legible to people who were not in it. That skill, applied to your own history in an interview, produces answers that sound excellent and contain no decisions — a smooth account of a thing that happened, with the candidate somewhere in the frame.
Experienced interviewers listen for exactly this, which is why so many PM questions are phrased to force a choice into the open: what you got wrong, what you cut, what you killed.
Influence versus attribution
The metric question is where PM CVs get tested, and the dynamic is different from other roles.
In engineering, a suspicious number is usually a measurement question. In product, the number is often real and the attribution is the problem. Activation genuinely went up 30%. Also: pricing changed, a growth team ran an onboarding experiment, a competitor had an outage, and the sales team started targeting a better-fit segment.
Everyone senior has lived this. So the follow-up is rarely a challenge — it is an invitation. “What else was going on at the time?”
Candidates who treat that as a threat defend the number. Candidates who treat it as the real question say what else was moving, then explain the part they can actually stand behind: the specific decision, the specific evidence, the specific thing that would not have happened otherwise. The second answer is more persuasive, and it is also just true.
This is worth working out before you are asked, because your CV probably states the number without the context. That is fine — a CV bullet has no room for confounders. But it means the number on the page is an opening, and you should know where it leads.
”What did you kill?”
Almost nobody prepares for this one and it separates candidates sharply.
Shipping features is the visible part of the job and the easy part to narrate. Stopping something — a project with a sponsor, a roadmap item people have already told customers about, a thing your own team is emotionally invested in — costs you political capital and produces no artefact to point at afterwards.
That is precisely why it signals seniority. If you have killed something, it belongs in your preparation even though it will never be a CV bullet, because it is evidence of judgement in a way that “launched X” is not.
Preparing from the document you sent
Go through your CV and, for every outcome number, write two things: the confounders you know about, and the one decision you would defend as causal. For every “led” or “owned”, name who else was in the room.
That exercise is uncomfortable and it is the whole preparation. Our free claim checker flags which lines an interviewer is most likely to open with, and the STAR guide covers the shape of the answers — though for PM interviews, the S and T matter less than the decision wedged inside the A.
Questions about the interview itself
- What questions are asked in a product manager interview?
- Typically four families: a product sense exercise (improve or design something), a metrics or analytical question, behavioural questions about decisions and conflict, and questions about your own shipped work. The last family is the one candidates prepare least and interviewers weight most, because it is the only part that cannot be rehearsed from a generic list.
- How do I answer 'tell me about a product you shipped'?
- Lead with the decision, not the feature. What was the choice, what did you know when you made it, what did you give up, and what happened. A description of the feature tells the interviewer what your team built; a description of the decision tells them what you contributed, which is the thing they are hiring.
- Do PM interviews still ask brainteasers and estimation questions?
- Estimation still appears, usually as a market-sizing or metrics question rather than a puzzle. Pure brainteasers have largely fallen out of use at companies with mature product functions. Prepare for the metrics version; it doubles as preparation for defending the numbers on your own CV.