Metrics question

What metrics would you use to measure the success of the Save feature at Facebook?

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 success metrics for a lower frequency, utility oriented feature within a larger app.

Clarify the goal of Save: help users return to content later, increasing long term value even if it is a low frequency action.
Propose a primary usage metric: percent of active users who save at least one item per month, and saves per active saver.
Propose a return metric: percent of saved items revisited within 30 days, since saving with no return has little value.
Propose an organization metric: use of collections or folders, signaling deeper engagement with the feature beyond a single save.
Add a guardrail: saving should not cannibalize immediate engagement, such as sharing or commenting, at the moment of discovery.
Tie metrics together into a funnel from save to return to repeat use, reviewed quarterly given Save's naturally lower frequency.

Separates the act of saving from the value of saving, correctly identifying return rate as the metric that actually matters.
Names a guardrail explicitly, cannibalization of immediate sharing or commenting, showing awareness of interaction tradeoffs.
Gives an illustrative number, e.g. assumes only 20 percent of saved items are ever revisited, signaling room for improvement.
Frames the metrics as a funnel, save to return to repeat use, rather than a flat list with no connective logic.

Measuring only save volume, which says nothing about whether the feature actually delivers value to users.
Ignoring the possibility that saving cannibalizes other, more immediately valuable engagement actions.
No metric tracking whether saved content is ever revisited.

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