Practice resources
A PM interview rubric you can actually use
Review a product case answer against five observable criteria. Includes scoring anchors, a weak answer, an improved answer and a partner practice routine.
Evaluate the reasoning, then the delivery
A confident answer can still avoid the difficult decision. Use this practice rubric to locate what is missing: a clear problem, evidence, an explicit trade-off, a measurable outcome or an executable next step.
This is an original Product Mastery practice rubric, not an employer scorecard or a hiring standard. It is designed for self-review and peer practice. Different cases can justify different choices.
Five criteria, three observable levels
| Criterion | 0 — missing | 1 — partial | 2 — explicit |
|---|---|---|---|
| Problem framing | Jumps straight to a feature | Names a user and a broad problem | Names a user, a specific task and a constraint |
| Evidence | Treats assumptions as facts | Cites a signal without its limits | Separates observations, assumptions and a test |
| Trade-offs | Lists everything as a priority | Chooses without an opportunity cost | Chooses, defers and names a reversal condition |
| Measurement | Uses an unbounded activity count | Names a metric without a window | Defines a denominator, window and guardrail |
| Execution | Ends with “align the team” | Suggests steps without ownership | Names an owner, a bounded test and a review point |
Case: activation is weak in a team notes tool
Synthetic scenario: 100 new team workspaces were created last month. Forty created a note; twelve shared a note with a teammate. Support heard that inviting a colleague is confusing, but the interviews covered only five teams. One engineer is available for a week. Decide what to do next.
Weak answer: “We should add AI summaries, improve onboarding and launch a referral programme. We can measure engagement and make the experience seamless.”
- The user and stalled task are unclear.
- The proposal exceeds the stated capacity.
- Engagement has no definition, denominator or review window.
An answer with visible judgment
“I would investigate first-time team creators who write a note but do not share it. The funnel suggests a gap between creation and sharing; it does not prove invitations cause it. I would ask five recent creators to share a note while we observe where they stop, and check that sharing events are logged correctly.
“If that confirms invitation friction, I would use the engineer’s week on a smaller invitation change and hold AI summaries. We would compare the proportion of new workspaces sharing a note within seven days, while watching failed invitations and support contacts. A named PM and engineer would review the small cohort after two weeks; if teams share but nobody returns, we would revisit whether sharing is the right value milestone.”
A reasonable review might score this answer 2 on each criterion. That means the reasoning is visible under this rubric, not that the proposed feature is proven correct. Another well-supported decision could earn the same score.
A 15-minute practice routine
Compare the content of the first and revised answer. Avoid treating the total out of ten as a verdict on your ability. Keep reviewer disagreements: they often reveal an unstated assumption worth making explicit.
- 3 minutes: read the case and write down assumptions.
- 4 minutes: answer aloud or in writing, without looking at the rubric.
- 3 minutes: a partner assigns a score and quotes one sentence as evidence for each criterion.
- 3 minutes: revise only the weakest criterion.
- 2 minutes: repeat the decision and explain what changed.
Common review mistakes
Product Mastery provides exercise-specific AI feedback in the signed-in product. The public rubric above is a self-review resource; it is not a transcript of a live learner result.
- Rewarding a fashionable framework even when it hides the decision.
- Penalising a reasonable alternative because it differs from the sample.
- Inventing data to sound certain.
- Treating a fluent AI-generated answer as evidence of your own practice.