Skip to content

Dating for people who actually listen

Your listening history tells a story. The artists you come back to at 2am. The deep cuts you've played 300 times. The genre rabbit holes that consumed an entire weekend. That is identity.

Try the demo

How it works for music heads

Connect your streaming. Spotify, Apple Music, Last.fm, SoundCloud, Bandcamp - wherever you listen. We pull your top artists, genres, listening depth, and audio features. Nothing is shared raw - only signals.

The algorithm understands that music exists in layers:

Niche overlap > mainstream overlap

Both swiping right because you both like Arctic Monkeys? That is a popularity 85 overlap. It barely moves the needle. Both deep into Igorrr? That is a signal.

Mainstream
Arctic Monkeys
Popularity 85 → NicheWeight 0.15
Niche
Igorrr
Popularity 18 → NicheWeight 0.82

But it goes further. The algorithm knows the difference between "heard one track for 90 seconds" and "1,200 minutes across 6 months, every album saved." Engagement depth and saved status turn a flat overlap into a real affinity score.

What gets scored

Shared artists weighted by how niche they are (Spotify popularity, Last.fm listener count, or platform-level frequency)
Shared genres - not just "rock" but the sub-genres: post-punk, math rock, shoegaze, breakcore
Listening depth - minutes listened per artist, not just "top 5"
Saved status - did you actually save it, or did it autoplay once?
Audio features - danceability, energy, valence, acousticness. Two people who gravitate toward the same sonic profile have something in common even if the artists don't overlap
Shared dislikes - genres you both skip or artists you've actively avoided. A mutual hatred of something popular is a bonding signal too

Your Wrapped is a dating profile

Except it actually means something.

"You both listen to black midi (popularity 24), you've both logged 400+ minutes on Black Country, New Road, and your audio feature profiles overlap 78% on energy and acousticness."

No vague "you both like music." Actual data. Actual niche.

The deep cuts matter most

Sharing a love for an artist with 12 listeners on Spotify contributes 9x more signal than sharing one with popularity 90. The algorithm is designed for people whose taste is their identity - not a checkbox.

Your crate-digging habit, your 3am Bandcamp purchases, your obscure SoundCloud follows - that is the data that finds you someone who actually gets it.

Integrations

Spotify
Top artists, genres, audio features, listening history, saved tracks
Apple Music
Play counts, library, loved tracks
Last.fm
Scrobble history, top artists/albums/tracks over time
SoundCloud
Likes, reposts, followed artists
Bandcamp
Purchases, collection, wishlist

All opt-in. All revocable. We show signals, never raw data.

Affinity Atlas is in development

No real matching is live yet. If you want to be the first to know when it launches - or you have ideas for how music matching should work - get in touch.

Stay in the loop