How many questions should a product manager interview have?

Five, asked of everyone, beats fifteen chosen on the day. Beyond about five you are trading depth for coverage, and depth is where the differences show — the second and third follow-up on one situation tells you more than a sixth topic.

Should I send the questions in advance?

For the situational ones, no — preparation is exactly what you are trying to see past. For anything about their own history, yes: making someone recall a specific project under pressure measures memory, not judgment, and the good answers get better with a night to think.

Do these work for a first product hire?

They work better. A first product manager has nobody above them to check the call, so what you are buying is entirely judgment under ambiguity — and every question here is about exactly that.

Product manager interview questions worth asking

By Sergio Gualda · 8 min read · Updated 14 August 2026

A question discriminates when a strong and a weak candidate give visibly different answers to it. Most standard product manager questions fail that test because the good answer is public and rehearsable. The ones that work put the candidate in a situation with no correct option — a request that contains its own answer, a stakeholder who will not move, a deadline that halves — and read what they reach for first.

What makes a question discriminate

A question is only useful if the answers spread. 'How do you prioritise?' does not spread: RICE, ICE and the rest are a search away, and a candidate who has never prioritised anything can describe them fluently.

The test is simple. Before you ask something, decide what a bad answer would sound like. If you cannot picture one — or if the bad answer is just a less polished version of the good one — the question is measuring presentation.

Seven that spread

Each of these has no correct answer, which is exactly why they work. What you are reading is what the person reaches for first.

Sales lost three deals over a missing feature and the CRO wants it next sprint. The data says churn is concentrated elsewhere. Write your reply to the CRO.
Estimate the value of a feature you know nothing about, out loud, right now. I care about the structure, not the number.
Tell me about something you killed. Who wanted it, and what did you say to them?
The deadline just halved. What goes, and who do you tell first?
Which single metric would you move, by how much, and what would you expect to get worse?
Tell me about a decision you now think was wrong. How did you find out?
Your engineer says the thing you scoped will take a quarter, not a sprint. You do not have a quarter. What now?

Four to stop asking

These are common, and each fails the spread test for its own reason.

'How do you prioritise?' — the frameworks are public and reciting one proves nothing about using one.
'How would you improve our product?' — measures how long they spent on your site last night, and quietly asks for free consulting.
'How many golf balls fit in a bus?' — measures comfort with a party trick. The correlation with product judgment is not established, and it filters on confidence.
'What is your biggest weakness?' — has one rehearsed answer, delivered by everyone, learned from the same posts.

How to read the answers

The pattern that matters is whether the candidate goes to the problem or to the process. Asked about a lost deal, weak answers describe how they would run a discovery cycle. Strong ones ask which deals, and whether the missing feature was the reason or the reason given.

The second pattern is whether they can be specific about a cost. Every real decision gives something up. A candidate who cannot name what theirs gave up has either not made many, or has not noticed.

Ask everyone the same ones

Unstructured interviews are among the weakest predictors in the hiring literature, and the reason is not that interviewers are careless — it is that when everyone gets different questions, nothing is comparable, and what you end up ranking is rapport.

Pick five, ask them in the same order to every candidate, and write down the answers before you discuss anyone. It costs nothing and it is the single largest improvement available to most hiring processes.

Frequently asked

How many questions should a product manager interview have?: Five, asked of everyone, beats fifteen chosen on the day. Beyond about five you are trading depth for coverage, and depth is where the differences show — the second and third follow-up on one situation tells you more than a sixth topic.
Should I send the questions in advance?: For the situational ones, no — preparation is exactly what you are trying to see past. For anything about their own history, yes: making someone recall a specific project under pressure measures memory, not judgment, and the good answers get better with a night to think.
Do these work for a first product hire?: They work better. A first product manager has nobody above them to check the call, so what you are buying is entirely judgment under ambiguity — and every question here is about exactly that.