Spotify Streaming History 2.0

The code can be found here.

Updated: January 26, 2026. Includes streaming data from January 2016 through December 2025.

If you're asking if I had the foresight to begin tracking my Spotify streaming history using last.fm in 2015 so that one day I could analyze myself, the answer is yes. And, today is that day.

Below, you can see my annual streams for the past decade.

In order to build genre, I scraped https://everynoise.com/. This defines genre at the artist level. I also added on some rules to overwrite the genre in cases such as remixes, which were classified as EDM. Christmas songs were also separated out into their own category by calling Spotify's API to pull down the names of songs on my Christmas playlist.

While this is the best data availabe for genre, a shortcoming is that it is defined at the artist level instead of by song. As an example, let's look at Lady Gaga, whose music spans multiple genres. While Lady Gaga gets classified as Dance Pop, it can easily be argued that songs such as Shallow or Die With a Smile fall under Rock and songs such as Aura or Alice fall under EDM.

Below, you can see how the percentage of streams by genre has changed over time by quarter. Christmas can be seen popping annually in the fourth quarter. Pop has lost share since a decade ago - mostly being taken over by the growth in Hopebeat/Indie and Dance Pop.

Below is every song I've listened to in the last decade, which is truly wild to think about. In the last decade, I've gone from a college student to studying abroad to my first job and apartment to homeowner. Through that all, this was the sountrack to that life that was happening.

Narrowing down to some more specific points in time, I have compiled a list of most played songs in a single day. These are the songs played obsessively over and over again. (And I know Born Ready should be higher, but I know I was studying abroad at the time, so the streams likely got split across two days.)

If we look into which artists were played the most in a single day, this is usually associated with an album drop. One quite impressive feat that I noticed here was Molly Sanden racking up 3 consecutive days that all made it into the top 25.

I also wanted to narrow my focus to look just at last year - 2025. Below is a dive into my top streaming songs of the year.

One thing I wanted to take a closer look at was time of day and if that had any correlation with the type of music I listen to.

Below, you can see the distribution of listening by genre and time of day, which for the most part lines up with my routine. There isn't much going on between midnight and 6 AM because I am asleep - or at least trying to be. This ramps up in the morning but drops off about lunchtime, which is when I might listen to a podcast. Then, it really explodes between 3PM to 7PM because this is the time I'm usually done with a podcast or at the gym. Finally, it falls off through the night as I come home and watch tv or something.

I wanted to take this one step further though and see if there was a correlation between genre and listening time. My hypotheis was that there may be more EDM around that 3PM to 7PM window because my gym playlist has quite a bit of EDM on it. Therefore, I built an expected value. If genre and time were completely independent of each other, then the actual listening should match the product of the average genre and average listening for that hour. Any deviation from that expected value would indicate the variables are related.

EDM clearly outperforms its expected value during my time at the gym.

On the other hand, Dance Pop almost perfectly matches its expected value all day.

Pop tails off and starts underperforming expectation as I get further into the workout and overperforms in the morning.

Both Hopebeat/Indie and Rock slighly outperform at that 5 PM mark, while Rock also outperforms in the afternoon.

Christmas extremely overperforms in the morning while underperforming in the afternoon. This is no shock to me. You try exercising to Have Yourself a Merry Little Christmas; it's not happening.

Finally, Other is all over the map, which is to be expected. The category is composed of a bunch of leftover genres such as synthwave, songs from entertainment media, etc.

In conclusion, was this project a little self-conceited? Yes. But really, this was a fun project to understand my own music tastes better and do some more webscaping and visualization work. And if you've made it this far, maybe you've been convinced to set up your own last.fm account so that you can track what you listen to as well. (You don't have to go crazy like me. You can just see your top songs and artist on the site.)

If you have a similar style of music and want to explore what I'm currently listening to, these are the songs currently in the cycle.