Metrics question

50% of users who RSVP yes to Facebook Events do not show up. What would you do?

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

Whether you can diagnose a gap between stated intent and actual behavior and target a fix at the real cause.

Define the metric precisely: no-show rate among yes-RSVPs, and confirm whether this is worse for certain event types.
Segment the no-shows: large public events versus small friend gatherings, since intent-to-attend differs a lot by context.
Generate hypotheses: RSVP yes used as a soft interest signal rather than a commitment, no reminder close to event time, or last-minute conflicts.
Check data support for each, such as comparing no-show rate for events with and without a day-of reminder sent.
Prioritize the reminder and friction hypothesis first, since it's the cheapest to test and likely explains a meaningful share.
Propose a fix: stronger day-of reminders and a lightweight re-confirmation prompt, and measure the change in no-show rate.

Frames RSVP yes as a soft signal of interest rather than assuming it always means firm commitment.
Segments by event type, since a 50 percent no-show rate likely hides very different causes for public versus close-friend events.
Proposes a cheap, testable fix, day-of reminders, before assuming a deeper redesign of RSVP is needed.
Gives an illustrative number, assuming events without a reminder show a materially higher no-show rate than those with one.

Assuming RSVP yes always means firm intent and treating every no-show as a broken promise to fix with penalties.
Proposing a heavy-handed fix, like penalizing users, before testing a lighter one like reminders.

Would you consider changing the RSVP options themselves, like adding a maybe with lower commitment framing?
How would you tell whether reminders alone closed most of the gap?