Metrics question

Zoom has integrated AI into their system, which includes automated notes, and since then, the number of meetings has reduced. What will 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

Tests diagnosing whether a usage decline caused by a feature is actually good or bad for the business, requiring reframing the problem.

Question the framing first: fewer meetings could mean the AI notes feature is working exactly as intended, reducing unnecessary meetings.
Segment meetings avoided because notes made them unnecessary, a good outcome, from meetings avoided due to dissatisfaction or churn risk, a bad one.
Check business impact: does reduced meeting volume correlate with reduced subscription usage or renewal risk, or is overall engagement stable.
Recommend shifting the north star from meeting count to value delivered per user, time saved or task completion, since the feature's own success makes meeting count outdated.
Define success as retention and renewal rate and NPS post-feature, not raw meeting count.

Explicitly reframes the premise, fewer meetings isn't automatically bad, rather than panicking at the metric drop.
Proposes checking renewal and retention as the metric that actually matters for Zoom's revenue model.
Draws a clear line between good-cause and bad-cause reductions in meeting volume.

Assuming fewer meetings is automatically bad and encouraging more meetings, contradicting the product's own value proposition.
Not checking whether retention or revenue is actually affected before reacting.

How would you distinguish meetings avoided because AI notes replaced them from meetings avoided due to churn risk?
What metric should replace meeting count as the real north star here?