Metrics question

What would you do if, all of a sudden, our web traffic dropped by 100%?

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

Incident-response diagnosis, distinguishing a real product problem from a technical outage.

Recognize a 100% drop almost certainly signals a technical outage or tracking failure, not a gradual behavior shift.
First check whether the site or service is actually down, versus just the analytics pipeline being broken.
If it's a real outage, escalate immediately to engineering and on-call rather than doing a product investigation.
If it's a tracking issue, verify with an alternate data source, e.g. server logs versus the analytics tool.
Once resolved, run a postmortem: root cause, time to detection, time to resolution, prevention steps.
Note who needs to be notified and how fast.

Immediately recognizes a 100% drop signals an outage or broken tracking, not a normal metric investigation.
Prioritizes checking site or service availability directly before doing any deeper analysis.
Cross-verifies with an independent data source to rule out an analytics pipeline failure specifically.
Proposes a postmortem afterward to prevent recurrence, showing process maturity beyond just fixing the issue.

Treating a 100% drop like a normal metric-decline investigation instead of recognizing it as an incident.
Not distinguishing between the service being down and the tracking being broken.

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