Your Spotify data has more secrets to reveal
At the end of every year, music fans eagerly wait for Spotify Wrapped to reveal their listening secrets. But you don’t have to wait until December to learn more about your music habits. In fact, you can use ChatGPT to create a much more personalized version of Spotify Wrapped that tells you exactly what you want to know.
Here’s how to launch your own Spotify Wrapped, whenever you want.
Request your Spotify data
The first step is to request your listening history from Spotify. The company allows you to request data for the past year or for the entire life of your account. If you just want the past year’s data, Spotify should send it to you in a few days. If you want lifetime data, recovery can (incredibly) take up to 30 days.
To request your data:
- Visit the Spotify website and log in with your account details
- Click on your profile picture in the upper right corner and select Account
- Scroll down and select the privacy setting
- On this screen you should find the two options to request your data
When the data is ready, Spotify will email you a link from where you can download the data folder.
Upload your data to ChatGPT
For this, you’ll need a ChatGPT Plus subscription ($20 per month), as the free tier doesn’t let you upload your own files for analysis.
Open the Spotify data folder and look for a file called “StreamingHistory_music_0.json”.
You can actually double-click this file and open it with any web browser. There you should find a long list of all the tracks you’ve listened to on Spotify. The data includes the precise time and date you played the song, the artist name, the song name, and the length of time you played the song in milliseconds.
Now start a new chat with ChatGPT Plus and drag this StreamingHistory file into the chat window or tap the paperclip icon and find the file on your computer.
Once downloaded, you can start asking ChatGPT to analyze the data in many different and interesting ways.
What to ask ChatGPT about your Spotify data
Find out when you listen to your favorite bands
You’ll already know your best artists and tracks of the year thanks to Spotify’s own data. However, you can dig deeper here.
For example, if you want to know how your listening habits have changed over time, you can ask: “Show me which artist I listened to the most each month of the year.” » If you use your account history data, you can request monthly data for each month since your account was opened, to get a much longer view of how your listening habits have changed ( or not).
Instead of a monthly breakdown, you can ask ChatGPT which artists you listen to the most each day of the week. To dig even deeper, you can even break it down by time of day.
For example, I asked ChatGPT to “Give me a list of the artists I listened to the most, on average, at each hour of the day.” So, for example: 00:00-00:59 Madness, 01:00-01:59 Oasis, etc. (It’s always good to give ChatGPT an example of the result you expect to reduce errors.)
In my case, Spotify returned the following data:
00:00-00:59: High Flying Birds by Noel Gallagher (5 pieces)
01:00-01:59: Noel Gallagher (3 pieces)
02:00-02:59: Alabama Shakes (1 reading)
03:00-03:59: Billy Bragg (1 piece)
04:00-04:59: The Sands (1 piece)
05:00-05:59: Irish Amber Leigh (1 piece)
06:00-06:59: Press (12 plays)
07:00-07:59: Gaz Coombes (11 pieces)
08:00-08:59: Unknown artist (17 pieces)
09:00-09:59: Fred again… (17 plays)
10:00-10:59 a.m.: U2 (17 listens)
11:00-11:59 a.m.: Paul Heaton (15 pieces)
I’ve shortened the list here, but it’s indicative of the kind of results you can get, with Fred’s more energetic tracks again usually going high when I’m at the gym, and the “dad rock” starting to make effect more. at lunch time.
You will notice in this list that “Unknown Artist” makes an appearance at 8am. It seems that when Spotify removes a track/artist from its collection, your data simply shows “unknown” in the artist and track name fields. To prevent strangers from clogging your results, you can ask ChatGPT to rerun the “exclude unknown artists” query.
Advanced Spotify Analysis
What more can you get from your data? Well, you can explore specific artists. For example: “Show the ten most played Radiohead songs. » Again, this type of request is more interesting with data over the entire lifespan rather than just one year.
The length of time you played the song also opens up some interesting possibilities. If you’re like me, you spend a lot of time listening to Spotify-generated playlists, skipping the tracks you don’t like. If you want to get a list of the artists you most often ignore, you can ask: “List the ten most played artists for whom the ‘ms played’ is less than 30,000.”
This will show you the artists you skipped the most in the first 30 seconds of playing the song, perhaps giving you a list of artists you can add to your Spotify blacklist to prevent their songs from playing in automatically generated playlists. (Unfortunately, this option to block specific artists still isn’t in Spotify’s desktop app.)
Finally, how about using your music history to create a quiz to test your musical knowledge? You can ask ChatGPT to “create a quiz for me based on my favorite artists and tracks.” Before you press Enter at this prompt, ask ChatGPT to act as the quiz master and let you enter your answers into the chat box, otherwise it will just spit out the questions and answers together and ruin your fun.