~/learn/Product Metrics · Trade-offs & launch decisions
Launch trade-offs & recommending a decision
Put numbers on trade-offs and give a clear ship / don't ship / iterate recommendation.
why this matters
Senior interviews end with 'would you launch it?' — a clear, quantified call stands out.
Senior product case rounds end with 'So, would you launch it?'. Results are rarely clean: the goal metric is up, but a guardrail dipped, or one segment got worse. The interviewer wants a clear recommendation with reasoning — not 'it depends'.
Convert trade-offs into the same units when you can. If a feature adds $2M a year in revenue but loses 0.3% of retained users, estimate what those users are worth over their lifetime and compare.
Decision framework
1. State the result: goal metric, guardrails, segments 2. Quantify trade-offs in shared units (revenue, users) 3. Consider long-term & strategy (trust, competition) 4. Decide: ship / don't ship / ship to a segment / iterate 5. Say what you'd monitor after launch
Gain: +$2.0M/yr ad revenue Loss: 0.3% of 4M users × $60 lifetime value ≈ $0.72M Net: ≈ +$1.3M → ship, with a retention alert
Interview trap
Don't ignore a negative segment. 'Overall positive, but new users retain 4% worse' might mean shipping to existing users only while fixing onboarding.
Pro tip
End every case with one clear sentence: 'I'd ship to existing users, hold back for new users while we fix onboarding, and watch day-7 retention weekly.'
key takeaways
- Put trade-offs in shared units.
- Don't ignore negative segments.
- Give one clear recommendation + monitoring plan.
check your understanding
Revenue +3% overall, but new-user day-7 retention −4%. Best recommendation?