Metrics question

How would you measure the success of Gmail?

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

Metric framework for a utility email product where value is largely invisible and habitual.

Define success as reliable, fast, spam free email that users trust enough to make their primary inbox.
Propose a reliability metric: email delivery success rate and time from send to delivery.
Propose a quality metric: spam false positive rate (legitimate email wrongly filtered) and false negative rate (spam reaching the inbox).
Propose an engagement metric: daily active users who check and act on email, and average time to inbox zero.
Propose a retention metric: percent of users who use Gmail as their primary email address across a year.
Tie these into a top line trust and reliability score, since Gmail's core value is being dependable, not entertaining.

Prioritizes reliability and spam accuracy as core metrics, correctly reflecting what actually matters for an email product's trust.
Separates false positive from false negative spam rates, since they represent very different user harms, missed email versus inbox clutter.
Gives an illustrative number, e.g. assumes targeting spam false positive rate under 0.1 percent as critical to trust.
Names primary email address usage as a long term retention proxy, capturing genuine habitual reliance on the product.

Naming only DAU or session count, which says nothing about reliability or trust, the core value of email.
Not distinguishing spam false positives from false negatives, which have very different user impact.
No metric tied to whether Gmail is the user's actual primary email address versus a secondary account.

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