How to Interview a PM

Most PM interviews tell you how someone thinks about product in theory. That's not the question. Here's the format that actually surfaces judgment.

The Setup

Before the interview, ask the candidate to send a one-page summary of a real problem they personally worked on. One hour. Any format. No template. A real problem, real ownership, real outcome.

The constraint does most of the work.

Forces prioritisation upfront — which problem do they choose, and why that one?

Live interviews favour confidence. One-pagers favour judgment.

Quieter candidates often outperform here. Hard to rehearse.

Strong signals What good looks like

✦

Sharp problem framing — not context Gets to the actual problem fast, skips the preamble

✦

First-person ownership "I decided" not "the team concluded"

✦

Options considered and rejected — with a reason Shows the thinking behind the call, not just the outcome

✦

Trade-offs consciously made, not just named

✦

Measurable outcome — or an honest admission of what wasn't tracked

Weak signals What to watch for

✗

Process with no actual decision anywhere in it "Discovery was run. Customers were spoken to." Who decided anything?

✗

Ownership always attributed to a team that can't be named

✗

No clear problem statement

✗

No trade-off surfaced — no evidence anything was given up

✗

Dense wall of text, every sentence pulling equally hard No hierarchy = no judgment about what matters

In the interview — the document becomes the script

Choice Why this problem over everything else you could have picked?

Rejection What did you consider and rule out — can you explain that quickly?

Shift What changed between your first read of the situation and how it ended up being solved?

Hindsight What would you do differently now?

No hypotheticals. Every question comes from the page. Hard to rehearse because it's grounded in something that actually happened.

What you're looking for by level

Junior

Can they own a delivery?

Do they understand what they learned? Can they articulate it clearly?

Mid-level

Did discovery actually happen?

Or is "we spoke to customers" doing a lot of work to hide that it didn't? Was there a trade-off that cost something real?

Senior+

Does it carry commercial weight?

Is there evidence the influence reached beyond the immediate product team?