An interactive story

We think we know
genre.
The data has an
opinion.

Six genres. ~33,000 songs. 21 years of music. Is genre just a sound, or a story?

Guess the genre

Listen to each preview, then guess the genre.

Sonic fingerprint

Mystery profile

Now playing

Song 1

Listen

Guess the genre

Track 1 of 5

How do you define genre?

— OSU College of Liberal Arts

— Franco Fabbri

— The New Yorker

Three definitions from three different fields — literary, academic, and journalistic. What they all agree on: genre is real, it's social, and it changes. You just heard it yourself.

Under the hood

Every genre has a sonic fingerprint

Those guesses weren't random — your ears were picking up on patterns embedded in thousands of songs. Here's what the data says each genre actually sounds like.

Use the feature tabs to re-rank genres by average audio traits. Hover a bar for the exact value, or click a genre to update the profile card.

The twist

The fingerprint is real. The hit isn't guaranteed.

Rap is wordy. EDM is danceable. Rock is loud. The patterns hold up across thousands of songs. But when we matched those fingerprints against streaming popularity, the correlation nearly vanishes.

Digging deeper

So what actually drives streaming success?

We tested song length, eight audio traits, and artist consistency against popularity. The results are messier — and more interesting — than you'd expect.

Behind the average

Behind every genre average is an artist making a bet.

Statistics describe genres. Artists define them — by deciding when to experiment, when to stay in their lane, and when to reinvent themselves entirely. Let's follow specific careers and watch those decisions play out year by year.

Artist journeys

How do artists evolve over time?

Pick an artist and step through their career year by year — then zoom out to see how the whole genre landscape shifted.

Recommended

Artists with 7+ years in our data

Zoom out

What one artist does, a generation follows.

Individual artists evolve. So do entire genres. Multiply those choices across thousands of artists and 21 years, and the sonic drift becomes undeniable — some traits change slowly, others shift all at once.

The big picture

How has music changed since 2000?

Pick a trait and a genre to watch the average shift year by year.

Low
High

Spotify playlist dataset · ~33,000 tracks · collected ~2020

What we learned

Three things the data taught us

01

Your ears were right all along

You can hear genre before you can name it. That instinct isn't imaginary — it's detecting real patterns in tempo, energy, and texture that consistently separate one genre from another across thousands of songs.

02

Popularity is harder to engineer than the sound

We tested every audio trait against streaming numbers. None predicted success very well. What does? Timing. Recognition. The cultural moment a song arrives in. The music is almost beside the point.

03

The genre you love already changed

It didn't happen overnight. Acousticness fell. Danceability climbed. The genre you grew up with sounds different now — shaped by millions of small artistic choices adding up over two decades.

Made with the Spotify playlist dataset · ~33,000 tracks · DSC 106