Metrics question

How would you set the goal for FB's Crisis Response feature?

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

Setting appropriate goals for a safety-critical, trust-sensitive feature rather than a pure growth feature.

Clarify the feature's mission: helping people mark themselves safe and find accurate, timely information during a crisis, not engagement maximization.
Define the primary goal: speed and coverage, i.e., how quickly and completely affected users can mark safe and access verified resources.
Define supporting metrics: percent of affected users reached by a crisis prompt, and time from crisis onset to prompt delivery.
Set a guardrail: accuracy of crisis detection and geofencing, since a false trigger causes real anxiety and erodes trust.
Avoid using engagement time as a success metric, since more time spent in a crisis tool likely reflects distress, not product value.
Define a secondary goal: information quality, like linking to verified official sources rather than unverified user posts.

Explicitly rejects engagement time as a success metric, showing sensitivity to the feature's emotional and safety context.
Names speed and coverage as the primary goal, appropriate for a time-critical safety tool.
Gives an illustrative benchmark, e.g. 90% of affected users receiving a prompt within 30 minutes of crisis onset, marked as an assumption.
Flags false-positive geofencing as a real risk, since incorrectly telling someone they're in a crisis zone damages trust.

Treating this like a normal growth feature and optimizing for engagement or session time.
Ignoring detection accuracy as a guardrail, which is critical for a safety feature.

Chapters of the AI PM course that teach what this question tests.