Metrics question

Pick a feature from a product of your choice & tell us how will you assess the success of that feature. Assume it is being launched now.

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

Ability to define a launch success framework for a self-chosen feature, showing metric tree thinking.

Pick a specific, well-known feature, such as Instagram's Close Friends list, to ground the answer in something concrete.
State the feature's goal: giving users a lower-pressure way to share with a trusted smaller audience, increasing sharing frequency.
Define a primary metric: adoption rate (percent of users who create a Close Friends list) and usage rate (stories shared to it weekly).
Define supporting metrics: overall stories shared per user (to see if it adds to, not cannibalizes, total sharing) and list size over time.
Set a guardrail: ensure the feature isn't just moving existing public shares into a smaller, more private space without net new sharing.
Define a launch timeline: track adoption in week 1 versus sustained usage at week 4, since novelty effects fade fast.

Grounds the answer in one real feature rather than staying abstract, showing concrete product knowledge.
Distinguishes adoption (tried it once) from sustained usage (still using it weekly), avoiding a novelty-effect trap.
Gives an illustrative target, e.g. 15% adoption in month one with 40% of adopters still active at week 4, marked as an assumption.
Flags cannibalization risk explicitly as a guardrail, showing awareness that a feature can succeed on its own metric while hurting the whole.

Staying too abstract and never naming a real feature to ground the answer.
Measuring only initial adoption without checking for sustained usage or cannibalization.

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