Inspiration

We love music and doom-scrolling on TikTok. Why not make a project that combines our 2 favorite things. We were also very interested in the relation between music platforms like Spotify and TikTok. How does a viral trend on TikTok affect the popularity of a song on Spotify?

What it does

Our project looks at which popularity boom comes first (TikTok or Spotify) and looks at information like time delay between the different spikes and its magnitude. We then compare the difference between whether the viral trend starts on TikTok or on Spotify. Does one affect the other faster or more significantly?

How we built it

We scraped data from Spotify and TikTok song specific data to track information like listeners and tiktoks made for the last 90 days which indicates the popularity of the song/trend.

Challenges we ran into

Figuring out how to use the data to describe and visualize a noticeable trend. We also had trouble finding a good set of data to work with that can show both TikTok and Spotify music trends simultaneously.

Accomplishments that we're proud of

Being able to work together on a project with people I've never work together with. Coming out with results and understanding the results although it doesn't give us the confidence to confirm our hypothesis. We developed parameters that tell us how influential a platform affects another. These are the correlation coefficient which equals the spike in popularity magnitude of the receiving platform divided by the spike in the initiating platform. Time delay is the number of days that one spike lead to a spike on another platform.

What we learned

We learned that TikTok seems to affect Spotify in a bigger magnitude while Spotify seems to affect TikTok faster. However, the standard deviation is too high for us to come to a confident conclusion.

What's next for tokapi

We can use many more platforms like Youtube, Soundcloud, Shazam. We can also work with more songs and work on viral trends from the past as well. That can be compared to current trends or to other trends depending on time period.

Built With

Share this project:

Updates