Spotify · Metrics · Mid · 45 to 60 minutes

Why are new Spotify users leaving after 2 songs?

Spotify is a streaming service where people listen to music and other audio on their phones and computers. Answer this product management case and get it graded line by line against a weighted rubric.

metric diagnosis
funnel analysis
hypothesis framing
prioritization
experiment design

The situation

You are a PM on Spotify's new user experience team. Your director asks for a root cause analysis after customer support, analytics, and app store reviews all point to a recent rise in uninstalls among people who signed up successfully but did not stick. Early event data suggests many new users stop listening after just one or two songs, yet installs remain healthy. Assume a recent sample of 100 new mobile users looks like this: 100 install, 90 sign up, 80 finish onboarding, 75 play a first song, 30 play a second song, 20 play a third song, and 10 return the next day. No one knows whether the issue is driven by onboarding, recommendations, app quality, acquisition quality, or something else. It also may not be uniform across countries, app versions, or acquisition cohorts. Leadership wants a clear diagnosis and a near term action plan before the next sprint, but there is only enough capacity for one focused investigation and a small test, not a broad redesign.

What to produce

Write a 1 to 2 page diagnosis and action memo.

Your answer must

State the scope you would use and the top clarifying questions or assumptions you need to make.
Describe the early user funnel from install to day 1 return, and name 5 to 7 metrics you would inspect first.
List at least 3 root cause hypotheses, each tied to a user segment or data cut and to evidence that would confirm or disprove it.
Recommend the first analysis slices you would use to separate cohort, region, and app release effects.
Propose one small next sprint test or intervention, with a success metric, a guardrail metric, and why you chose it over other options.

Constraints

Do not change Spotify's music catalog, licensing, or subscription pricing.
You have one sprint for analysis and at most one low to medium effort mobile app test.
Use only data and research a product team could gather quickly, such as product analytics, crash and performance logs, support signals, app store reviews, and a few user interviews.

Format

Use short sections with bullets and include a simple funnel table.

How Spotify actually played it

An RCA walk-through for a rise in Spotify uninstalls, covering clarifying questions, hypotheses by segment and a plan to test them.

Uninstall is a lagging signal, so find leading ones
Separate release, region and cohort effects
Frame hypotheses before opening dashboards

Built from Spotify users uninstalling the app: Root Cause Analysis case by Sreerag Panat.