The rumors about so-called "ghost artists" and the term "Perfect Fit Content" (PFC) used internally by Spotify were already uncovered some time ago in Liz Pelly's book Mood Machine. What I've been missing since then is quantifiable evidence about how common these artists actually are in major playlists. To shed some light on this, I've spent the last few days building a tool with Claude Code to identify the "ghosts". I'm currently cross-checking the results with Brian and wanted to share them with you.
Full methodology and results are in the HTML file below. 👇🏼
If you know of other suspicious playlists you'd like me to run through the tool, or if you have tips and ideas on how to improve the detection methodology, I'd love to hear your feedback. 🎸