Metrics question

Glean has strong pilot sign-up and early usage, but week-4 retention is weak. How would you diagnose whether the root cause is poor onboarding, weak core value, insufficient off-surface engagement, or failure to expand from an individual user to a team habit? Describe the funnel cuts, cohorts, qualitative inputs, and first 3 experiments you would run.

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Structured retention diagnosis across multiple plausible causes, using funnel and cohort data plus qualitative input.

Cut retention by first week behavior: compare users who completed a first meaningful search or query against those who did not, to isolate an onboarding activation gap.
Cut by usage surface, checking whether users engaging only inside the app churn more than those using it through Slack or browser extension integrations, pointing to an off surface engagement gap.
Cut by team density, checking whether solo users within a team churn more than users on teams with several active colleagues, which would point to a missing team habit effect.
Layer in qualitative interviews with a sample of week two churners to hear directly whether they never found value, forgot the tool existed, or found it useful but not habitual.
Based on the strongest signal, design three experiments, for example a stronger first session activation flow, a proactive off surface integration prompt, and a team visibility nudge showing colleague activity.
Set week four retention itself, cut by these same segments, as the metric to validate whether each experiment closes the specific gap it targets.

Runs multiple distinct diagnostic cuts, activation, surface, and team density, rather than guessing at one cause.
Uses direct qualitative interviews with churners to validate the quantitative hypothesis.
Proposes three experiments each mapped to a specific diagnosed cause, not generic retention tactics.
Names team density as a distinct hypothesis, reflecting a team habit dependent value proposition.

Assuming one single cause, usually onboarding, without checking off surface engagement or team effects.
Proposing experiments before doing any qualitative diagnosis of why users are not returning.

Which of these three causes would you bet on before seeing the data, and why?
How would you distinguish a core value problem from an onboarding problem in the data alone?