Metrics question

You're the product manager at Spotify. Build metrics.

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

Building a metric tree from scratch for a subscription streaming product: can you connect listening behavior to Spotify's actual business model.

State the business goal: grow paid subscription revenue while keeping free users engaged enough to eventually convert.
Set the north star as hours streamed per active user, since it correlates with both engagement and ad exposure or subscription value delivered.
Break it into a funnel: free user activation (first week listening), free to paid conversion rate, and paid subscriber retention.
Add creator or content supply metrics: catalog engagement diversity, since heavy reliance on a few hit songs is a business risk, unlike broad catalog usage.
Add a guardrail: churn rate for paid subscribers, since acquisition gains mean little if paid users leave quickly.
Define the top metric for leadership as net new paid subscribers combined with retention, since that most directly reflects revenue health.

Connects the metric tree to Spotify's actual two sided business, free to paid conversion plus retention, not just raw listening hours.
Adds a catalog diversity guardrail, a nuance specific to a content platform that depends on broad supply, not just top hits.
Names a clear top line metric for leadership instead of leaving the tree flat with no hierarchy.

Measuring only listening hours without connecting it to the subscription conversion funnel.
Ignoring paid subscriber churn, which can offset any gains in new subscriber acquisition.

How would you weigh podcast metrics differently from music metrics in this tree?
What would you do if free user engagement is strong but conversion to paid stays flat?