Spotify wants a partner who can work at the intersection of algorithmic recommendation and human editorial curation. The brief isn't to replace the editorial team with AI — it's to make the editorial team's cultural intelligence legible to the algorithm, and the algorithm's signals legible to the editorial team. They want someone who understands taste as a system.
94% MatchThe recommendation engine knows what people play. It doesn't know why certain things matter at a cultural moment — why a song lands differently in October than in May, why a genre becomes a scene, why an artist crosses from cult to mainstream before the data reflects it. That's editorial intelligence.
The question is how to make that intelligence systematic — teachable to the algorithm without flattening it into signals. That's a cultural problem, not a data problem.
taste is a system. we've been building it.A daily AI cultural briefing that surfaces what's shifting in culture before it becomes obvious. Built on 150+ feeds, runs daily, automated. Demonstrates AI-powered editorial intelligence at scale — exactly the infrastructure Spotify's algorithm needs access to.
A cultural intelligence system for senior strategists. Brief-driven curation, not algorithm-driven. Demonstrates the human editorial layer that Spotify is missing — the structured capture of cultural judgement that can be passed to a system.
295 books, AI-curated, with voice takes. Consistent editorial voice across a large corpus. Demonstrates taste as a curatable system — the same challenge Spotify faces when building editorial coherence across genres, moods, and moments.
151 feeds, 154 brands, AI-powered cultural signal extraction. The infrastructure that makes editorial intuition systematic. This is the engine version of what Spotify needs to build between its editorial team and its recommendation architecture.
Map the cultural signals that Spotify's editorial team uses intuitively but can't yet pass to the algorithm. Interviews with playlist editors across 5 key genres. Identify the 12–15 signals that most reliably predict cultural relevance before streams reflect it. Deliver a signal taxonomy, not a strategy deck.
Build the system that makes those signals legible to the data team. Not replacing editorial judgement — encoding it. A structured brief format that editorial teams fill per playlist, per mood, per moment, that the algorithm can read alongside stream data. Light tooling, existing workflow integration.
Deploy, measure, iterate. Does encoding editorial signals before a release improve recommendation performance 4 weeks after? Test on 3 markets, 3 genres. Build the evidence base for scaling across the full editorial function. Monthly review with editorial and data teams together.
Built The Pattern (daily cultural briefing), Trove (cultural intelligence for strategists), Cultural Terminal (154-brand signal extraction). Former Contagious — the cultural intelligence agency — MediaMonks, R/GA. 15+ years at the intersection of culture, brand, and technology. The person who has been building the exact capability Spotify needs — just not inside Spotify yet.
For an 18-month engagement, we'd bring in a music culture specialist from our network for the genre-specific signal mapping. The system design and AI architecture stays with us.
Give us one Spotify editorial playlist — your choice of genre and moment. We'll return a cultural signal map: the 8–10 cultural factors that made those tracks land together, expressed as structured signals the algorithm can read. It takes us three days. It costs you nothing. If it changes how you think about the editorial–algorithm relationship, we've earned the conversation.