💾 Our takeaway:

Editorial curation still shows a "popularity bias", making it difficult for new artists to break through and increasing barriers to entry.

It raises regulatory questions regarding market concentration and cultural diversity in the recorded music industry 🎶

However, established artists start with more listeners on average, so the same proportional boost means much higher absolute gains (in both streams and revenue).

Using an event-study design (robust to staggered adoption), we estimate how streams change before and after playlist inclusion or drop.

Both groups see similar percentage gains (resulting from inclusion) and losses (resulting from drop).

When emerging artists do make it onto playlists, they’re typically:

• Placed lower in the track order
• Featured on less popular playlists
• Removed sooner than established artists

We define emerging artists as those with <2 years of career seniority.

Across all curators in France (Spotify + Major Labels), their share of playlisted tracks has declined since 2021.

💡Key findings:

• Emerging artists’ presence in editorial playlists is declining over time
• When featured, they get less visibility (lower rank, shorter stays)
• Playlist inclusion boosts streams similarly for everyone (in proportion)...
• ...but established artists win much more in absolute terms

😶

Purée un bureau à 4 ordis c’est mieux que la rolex a 50 ans 😎

Thanks, lots of sleepless nights to find it 😅

Not my field but perfect score 💯 for the title

Thank you!

...a beneficial effect for the regulator, who could use this lever to promote cultural diversity (via the discovery of new tracks through demand); but a more contrasting effect for platforms depending on their royalties distribution system (“market-centric” vs. “user-centric”)

These results could have important implications for the music industry 💽: a beneficial effect for artists and record companies thanks to the gain in engagement; ...

✅ Similar effects can be found in terms of feedback on the recommended tracks: higher like rate, lesser dislike rate, etc.

This effect is persistent during the firsts 30 tracks, then it decreases to reach the level of the other groups