What your Spotify says about your compatibility

Music as a compatibility signal

When you meet someone and discover you both love the same obscure band, something clicks. It is not just shared taste. It feels like evidence of a deeper alignment - similar values, similar way of seeing the world, similar emotional landscape. That feeling is not irrational. It is backed by research.

A 2011 study published in Personality and Social Psychology Bulletin found that when people share similar tastes in music, it often means they share similar values - which in turn makes them more likely to be attracted to each other and form strong social bonds. Music preference is not just a hobby. It is a signal.

Research from the University of Cambridge went further, finding that musical preferences unite personalities across the globe - the relationship between music taste and personality holds across cultures, ages, and geographies. A jazz listener in Tokyo and a jazz listener in London are more psychologically similar to each other than either is to their neighbour who listens to heavy metal.

This is why Affinity Atlas treats music as one of its highest-signal data sources. Not because music is inherently important to dating, but because what you actually listen to reveals who you actually are.

The Big Five connection

The most robust finding in music-personality research is the connection between listening habits and the Big Five personality traits: Openness, Conscientiousness, Extraversion, Agreeableness, and Emotional Stability (sometimes called Neuroticism).

Spotify's own research team found that music taste can detect personality traits - particularly Emotional Stability and Conscientiousness - with greater accuracy than previous methods using artist likes on social media. Their study showed significant correlations between listening patterns and personality across multiple variable categories: derived features, genre preferences, and mood preferences.

Verywell Mind's review of the literature summarised the broad patterns:

• Openness to experience correlates with preferences for complex, unconventional music - jazz, classical, world music, experimental genres
• Extraversion correlates with preferences for upbeat, rhythmic, social music - pop, dance, hip-hop
• Agreeableness correlates with softer, more emotionally accessible music
• Conscientiousness shows links to structured, predictable music with clear patterns
• People who enjoy a wide variety of music tend to be more extraverted, agreeable, and conscientious overall

Interestingly, a twin study published in PMC found that the link between personality and musical sensibility has a genetic component - musical sensibility has a heritability of 64%, and personality accounts for roughly 40-50% of variance. This is not just cultural conditioning. There is something genuinely deep about the connection between who you are and what you listen to.

A 2024 meta-review reinforced this: individuals who enjoy classical music tend to be more open-minded and creative; pop listeners are typically more extraverted and sociable; heavy metal fans lean towards adventurousness; jazz listeners tend towards emotional stability.

What Spotify actually exposes

When a user connects Spotify to Affinity Atlas via OAuth, the API provides several categories of data (with the user's explicit consent):

• Top Artists - your most-listened artists across short, medium, and long-term windows
• Top Tracks - your most-played songs across the same time windows
• Saved Library - albums, playlists, and saved tracks
• Audio Features - Spotify's own analysis of each track: danceability, energy, valence (positivity), acousticness, instrumentalness, speechiness, tempo, and key
• Genre Tags - genre classifications for each artist, often highly specific ("shoegaze", "vapor soul", "Indonesian indie")

The audio features are particularly valuable because they provide a quantitative signature of listening behaviour. Two users might not share a single artist, but if their top tracks cluster around similar energy, valence, and acousticness profiles, they are listening to music that feels the same - which, per the Big Five research, means they likely share personality traits.

Why niche overlap matters more

This is where Affinity Atlas's niche weighting system becomes especially powerful in the music domain.

If two people both listen to Taylor Swift, that tells you almost nothing. She has 90+ million monthly listeners. Liking Taylor Swift is a near-universal signal - it does not distinguish you from the general population. The specificity of the overlap is what matters.

The intuition is simple: the more obscure the shared taste, the more it reveals about genuine compatibility. Two people who both listen to Mamaleek - a experimental black metal band with 12,000 monthly listeners - almost certainly have a deep, specific overlap in aesthetic sensibility that goes far beyond music. They likely share values around artistic experimentation, emotional intensity, and counter-cultural identity.

This is not speculation. It is the same principle that powers collaborative filtering in recommendation systems, applied to human compatibility instead of product recommendations. As we detailed in How Niche Weighting Actually Works, the mathematical framework uses inverse popularity weighting to ensure that rare overlaps are valued exponentially more than common ones.

The Affinity Atlas music pipeline

Here is how Spotify data flows through the Affinity Atlas system, from raw API response to compatibility signal:

1. Data ingestion

When a user connects Spotify, the identity stitching layer pulls top artists, top tracks, and audio features. Data is normalised and deduplicated against existing catalogue entries. The user's listening profile is stored as a structured document, not raw API responses - we keep what we need and discard what we do not, per the privacy by design principles.

2. Feature extraction

From the raw data, the system extracts three types of features:

• Artist-level features: Genre tags, popularity tier, and listening intensity (how much of this artist's catalogue has the user explored)
• Audio profile: Mean and variance of audio features across top tracks, creating a "sonic fingerprint" - are they drawn to high-energy, danceable tracks or low-energy, acoustic ones?
• Discovery behaviour: What proportion of their listening is mainstream vs. niche? How frequently do they explore new artists? This is itself a personality signal (high Openness correlates with more diverse listening)

3. Niche weighting

Each shared artist or genre between two users is scored using inverse popularity weighting. Shared niche artists contribute exponentially more to the compatibility score than shared mainstream artists. The maths are detailed in the niche weighting deep dive, but the principle is straightforward: rare overlap = strong signal.

4. Cross-domain integration

Music data does not exist in isolation. It is combined with signals from other connected platforms - gaming habits from Steam, reading tastes from Goodreads, film preferences from Letterboxd. A user whose music, film, and book tastes all point towards dark, experimental, emotionally complex content has a much clearer personality profile than music alone would provide. The transparent algorithm shows users exactly how each signal contributed to their match score.

Honest limitations

Transparency means being honest about what music data cannot tell you. Here are the genuine limitations:

Context blindness

Spotify data does not distinguish between music you love and music you tolerate. The workout playlist, the focus playlist, the "my partner played this in the car" playlist - they all count the same. Audio features help somewhat (if your top tracks span wildly different profiles, you probably have contextual listening habits), but the signal is noisier than it would be if we knew why you listened.

The personality proxy is imperfect

A recent meta-analysis noted that while personality and music preference are consistently correlated, personality itself "explains little variance in music preferences." The relationship is real but modest. Music taste is influenced by culture, age, peer group, mood, and dozens of other factors besides personality. Affinity Atlas treats music as one signal among many, not as a definitive personality test.

Music matching beyond a single platform

Spotify is one of several supported music platforms. Affinity Atlas is designed to ingest listening data from multiple streaming services - including Apple Music, YouTube Music, Last.fm, and others - so that music compatibility is not gated by which app you happen to use. Users who do not connect any music platform are matched on their other connected data sources, and the system does not penalise them for the absence.

Shared taste is not shared values

Two people who both love death metal might have completely incompatible values, lifestyles, and relationship goals. Music taste is a signal, not a guarantee. Affinity Atlas uses it as one input to a multi-dimensional compatibility score - it raises the probability of compatibility, but it does not determine it alone.