The framing — what Discover Weekly actually does for the business:

Direct engagement: users listen to the playlist.
Retention driver: weekly 'new music' keeps users coming back on a Monday.
Discovery engine: users find new artists/songs they like and save to their libraries.
Word-of-mouth acquisition: shareability (people mention Discover Weekly publicly, driving signups).
Free-to-premium conversion: users who engage with DW may convert to premium at higher rates.

Each has a different measurement approach.

The rigorous measurement:

1. Holdout test. The only way to measure the net impact is to hold out a cohort that doesn't get Discover Weekly. Compare:

Weekly app opens.
Retention at 4, 12, 26 weeks.
Premium conversion.
Library growth.

This is philosophically hard — Spotify can't really take DW away from current users, but they can hold it back from new users during onboarding. Or A/B test its prominence.

2. Proxy: dosage response. Instead of a clean holdout, look at within-user variation:

Users who engage with DW more (listen to 5+ songs / week) vs. less (0-1 songs).
Control for pre-DW engagement level (matched cohort).
Post-matching, measure downstream retention and premium conversion.

This isn't causal but is directionally informative.

3. Natural experiment. DW rolled out to countries at different times. Compare retention trajectories in countries pre- and post-rollout vs. unchanged countries.

4. Discovery outcome metric. Count: new artists saved to library after being heard via DW. Proxy for 'DW gave me music I now like.'

What the typical result looks like (publicly: Spotify reported DW drives significant engagement lift):

"Discover Weekly drives ~30 min/week of additional listening per engaged user. Retention among users with 3+ weeks of DW engagement is ~8 pp higher than matched non-engagers. Net: DW is plausibly responsible for 10-15% of overall premium retention."

Guardrails:

Cannibalization of curated/editorial playlists. Does DW pull listening away from Today's Top Hits, etc.? Likely some; the question is whether DW monetizes better.
Artist economics. Does DW concentrate streams on a few artists, or distribute more broadly?
Library growth saturation. Do users eventually accumulate so much DW saved content that they don't need DW anymore?

What senior candidates add:

Reference the public Spotify engineering blog posts on Discover Weekly — they've been transparent about the impact.
Distinguish the product feature effect (DW itself) from the platform effect (recommendation culture Spotify has built; DW is one expression of it). The latter is much bigger but harder to measure.
Consider retention as the true north star for recommendation features. Engagement is a proxy; the business asset is a user who keeps paying.

Measure the business impact of Spotify's Discover Weekly.

Draft your answer

Stuck? Peek at a hint

More case studies questions