Metrics question

How do you measure the success of Google Search?

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 build a metric framework for a core, mature product where 'success' has multiple dimensions.

Define what success means for search: users find what they need quickly and trust the results enough to return.
Propose a primary metric: query success rate, for example searches with no reformulation or abandonment within the session.
Propose a speed metric: time to first meaningful result click, since latency directly affects satisfaction.
Propose a quality metric: click through rate on top results combined with human rated relevance scores.
Propose a business guardrail: ad revenue per search should not be optimized at the expense of organic result quality.
Tie these together into a top line health score reviewed regularly, rather than relying on one single number.

Separates success into distinct dimensions, speed, relevance, and satisfaction, instead of naming one vague metric like 'DAU'.
Names a specific proxy for quality, query reformulation rate, which is a well known real search industry signal.
Gives an illustrative number, e.g. assumes reducing reformulation rate from 20 percent to 15 percent as a success bar.
States the tension between ad revenue and organic quality explicitly, showing awareness of Google's actual business model conflict.

Naming only DAU or query volume, which says nothing about whether users are actually satisfied with results.
Ignoring the tension between ad monetization and organic search quality.
Proposing metrics with no way to actually measure relevance or satisfaction.

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