Diagnose a 20 percent drop in activation for Music

Problem Statement Description

Product context: Spotify is an audio streaming company; its products include music, podcasts, audiobooks, playlists, recommendations, creator tools, subscriptions, and ads.

Spotify has observed a 20% drop in activation for Music among free listeners. In this RCA interview, you are asked to diagnose what may be driving the decline and how you would structure the investigation before recommending any mitigation.

Assume “activation” refers to a meaningful early user milestone for free Music listeners, such as reaching a first successful listening session, playing a certain number of tracks, saving or following content, or returning after an initial session. You should clarify the exact definition, timeframe, baseline, and affected population before forming hypotheses.

Your investigation should reflect Spotify’s product context: global audio streaming, personalized discovery, ad-supported listening, onboarding flows, music search and playback, playlist recommendations, and competition for listening time from other audio and entertainment products.

The experience should consider:

- How to frame the anomaly: size of drop, start date, duration, seasonality, geography, platform, app version, acquisition channel, and listener cohort.

- Whether the activation metric itself changed due to instrumentation, logging, experiment bucketing, attribution, or definition changes.

- Segmentation across new vs. returning free listeners, iOS vs. Android vs. web, markets, languages, device types, and network conditions.

- Product funnel steps leading to activation, including signup/login, onboarding, permission questions, search, recommendations, playback start, ads, and first-session completion.

- Hypotheses across product changes, recommendation quality, catalog availability, ad load, performance issues, account creation friction, and external competitive or seasonal factors.

- Evidence you would seek from dashboards, event logs, experiments, release notes, customer support tickets, social feedback, and qualitative user signals.

- Immediate mitigation options, including rollback, experiment pause, targeted fixes, alerting, and stakeholder communication.

- Prevention mechanisms such as monitoring, metric guardrails, release checks, and clearer ownership of activation health.

The goal is to demonstrate a structured RCA approach: validate the drop, isolate where and for whom it happened, generate testable hypotheses, prioritize evidence, and outline how Spotify should respond without jumping prematurely to a single cause.