Understanding Why Spotify Fails as a Music Discovery Platform
Interviews and survey data on why people treat Spotify as a music player, not a discovery engine
✳The problem they went after
Users describe Spotify as the place music lands after they find it somewhere else. In Shaswat's research, 58 percent of recently liked songs were first heard on Instagram Reels while Spotify accounted for 8 percent of new discoveries. Trust in recommendations is thin: 41 percent say what plays matches their mood only sometimes, and 25 percent say Spotify never surfaces a song they end up liking.
✳What they actually did
Three hypotheses went into the research and came out graded differently. The cluttered, repetitive home page idea was only partly validated: the repetition is real, but it shows up as stale recommendations and playlist fatigue rather than a layout problem. Weak recommendations for discovery-driven listeners was confirmed in all five interviews. The trust gap held up too, with users leaning on familiar artists and outside platforms. The sharpest finding is the coping loop: 50 percent skip until something fits, 75 percent retreat to their own playlists, 66 percent search manually.
✳What you can take from it
Worth stealing
- Three hypotheses stated up front, then graded honestly
- 58 percent of liked songs first found on Reels
- Coping loop: skip, retreat to playlists, search manually
- Partial validation reframes clutter as recommendation staleness
Read this if
Learning to write hypotheses that research can actually disprove
The transferable bitEach hypothesis was written so research could knock it down, and the partial validation is the most valuable result of the three, because it moves the problem statement from home page clutter to mood-aware, context-sensitive recommendation before a single solution gets drawn.
✳The submission itself
Open the PDF →Reading on a phone? The embedded viewer is cramped — open the PDF in a new tab instead.
