Metrics question

What is the north star metric for Yelp?

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

Whether you can pick a north star metric that actually reflects the two-sided value Yelp delivers, not a vanity number.

Clarify Yelp's core value: helping users find trustworthy local businesses, and helping businesses get discovered by relevant customers.
Consider candidate metrics: monthly active users, reviews written, or searches leading to a business action like a call or visit.
Pick the metric that best reflects value delivered on both sides: searches or page views that result in a user action, like calling, visiting the website, or getting directions.
Explain why raw review count or MAU alone is insufficient, since it does not confirm the recommendation was actually useful.
Add supporting metrics: review quality and recency, since stale reviews reduce the platform's usefulness over time.
State the guardrail: review authenticity and fake review rate, since trust is the core asset Yelp depends on.

Picks a metric tied to actual usefulness, an action taken after a search, rather than a vanity metric like total reviews.
Explains explicitly why simpler metrics like MAU or review count fall short of capturing two-sided value.
Adds a trust guardrail, fake review or spam rate, appropriate for a platform whose core value depends on trustworthy content.

Picking a purely one-sided metric like review count with no tie to actual user or business value.
No guardrail around review authenticity, a core trust risk for Yelp.
Choosing MAU as the north star, which does not reflect whether the platform actually helped anyone.

How would you detect and handle fake reviews affecting this metric?
How would you weigh business-side satisfaction against user-side satisfaction in this metric?