Hiring rubric for a senior product manager.
Score candidates on what actually matters — weights calibrated to real hiring outcomes, with ten questions that probe each dimension.
Send this to your panel. Anyone with the link sees the same rubric.
The rubric.
Weights sum to 100. Calibrated across small-team hiring panels.
Ten questions to ask.
Each includes what a strong answer contains and what a weak answer sounds like.
-
01communication medium · launch
Write a tweet announcing a new feature. What's your editing process?
Strong answerCuts 50% of words. Leads with user benefit. Avoids buzzwords. Knows who they're writing for. Has colleagues read it.Weak answerFirst draft is the final. Lots of adjectives. -
02product hard · strategy
Your product has product-market fit in one segment. CEO wants to expand to a new one. How do you think about whether to do it?
Strong answerEvaluates whether current PMF transfers. Cost of serving new segment. Risk to existing segment's experience. Would talk to 10 prospects first. Discusses focus as a strategic choice.Weak answer'Sure, let's do it.' Or 'no, stay focused' — either without analysis. -
03product medium · failure
Tell me about a feature you shipped that failed. What did you learn?
Strong answerSpecific feature. Specific failure mode (low adoption, negative metric, user complaint). Clear lesson applied to next shipment.Weak answer'Everything worked.' Or vague 'we iterated.' -
04culture medium · feedback
A designer pushes back hard on a decision you made. How do you respond?
Strong answerCurious, not defensive. Asks what they're seeing that you're not. Separates their reasoning from their emotion. Changes your mind if they're right. Explains clearly if they're not.Weak answerEither steamrolls or caves. Either way, no real conversation. -
05experience baseline · resume-probe
Your resume says you grew product X's usage by Y%. What did YOU do vs what the team did?
Strong answerClear boundary between individual contribution + team. Credits specific colleagues. Knows why certain decisions were made even those they didn't make.Weak answerClaims credit for team output. Vague about own role. -
06product medium · prioritization
You have 10 feature requests from enterprise customers. How do you prioritize?
Strong answerFrameworks (RICE, ICE) used sanely, not dogmatically. Asks about strategic fit, not just loudest customer. Distinguishes 'painful' from 'asked for.'Weak answerWhoever yelled loudest or paid most. Or a Jira-style priority queue with no strategy. -
07stakeholder medium · eng-disagreement
Engineering tells you a feature you need will take 6 months. You think it should take 6 weeks. What do you do?
Strong answerAsks why. Doesn't assume engineering is gaming estimates or they're being unreasonable. Breaks down the 6 months into components. Proposes phased shipping. Respects engineering judgment while pushing back.Weak answerEither accepts silently or escalates aggressively. No curiosity. -
08situational medium · data-conflict
Your quant analyst says users don't want feature X. Your best customer says they do. Who do you believe?
Strong answerNeither blindly. Investigates — is it the same feature they're describing? Is the quant measuring the right thing? Talks to 5 users to triangulate.Weak answer'Data wins.' Or 'the customer is always right.' Either is wrong. -
09product medium · metrics
You're launching a new feature. What metric tells you it's succeeding?
Strong answerLeads with user-facing value (not just 'DAU went up'). Distinguishes leading from lagging indicators. Guardrails to catch unintended harm.Weak answer'Users will love it.' Or picks a vanity metric. -
10stakeholder medium · sales
Sales commits to a feature with an enterprise prospect that's not on your roadmap. How do you handle?
Strong answerDebriefs with sales to understand real need vs stated need. Evaluates fit with strategy. Clear yes-with-conditions or no-with-alternative. Doesn't punish sales for closing deals.Weak answerEither adds it silently or refuses without engaging.
Let Lucida run round one against this rubric.
Candidates take an adaptive interview, scored live against the rubric above. You get a shortlist with cited evidence and a booked calendar for the humans who clear the bar.