Metrics question

Traffic on the New York Times homepage has decreased by 15%. How would you identify the cause?

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

Root cause diagnosis of a metric drop: structuring a hypothesis tree across internal and external causes.

Confirm the drop is real: check measurement or tracking errors first, like a broken analytics tag or timezone bug.
Segment the drop by traffic source, such as direct, search, social, and referral, to isolate where the decline concentrates.
Check for internal changes, like a recent homepage redesign, paywall change, or site outage around the same period.
Check external factors, like an algorithm change from Google or a seasonal news cycle dip.
Prioritize the hypothesis with the strongest signal and propose a fix or further investigation.

Rules out a tracking bug first, since a 15% drop is sometimes a measurement artifact, not a real user behavior change.
Segments by channel and finds, for example, that the drop concentrates in search traffic, pointing to an SEO or algorithm issue.
Proposes a specific next step, like checking Google Search Console for ranking changes, rather than a vague investigate further answer.

Jumping to a root cause without first ruling out tracking or measurement issues.
Analyzing total traffic without segmenting by source, which hides where the real problem is.

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