Maxine Hsu, Shreena Thakker, Amanda Tsai, Zoey Tsai
The topic we chose is user churn on Spotify. The question we want to answer is:
Q. Based on a set of features associated with a specific user, will the user churn?
A user is considered “churned” if the user no longer actively uses the platform. To answer this question, we will analyze user data such as listening habits, frequency of app usage, and other engagement metrics to predict whether a user will remain active on the platform.
The reason we chose this topic is because user retention is a challenge that many businesses struggle with. In general, it costs more to acquire new users than to keep existing ones. By answering this question, companies like Spotify can gain an understanding of the factors related to user churn and then develop business strategies to reduce it, which will increase profit in the long run. Understanding why users churn will also help companies identify weaknesses in their product and where resources should be directed to improve it.
- Download the CMSC320_Final_Project.ipynb
- Open the notebook in Google Colab or Jupyter Notebooks
- Run the notebook to see the results.
You can also see the project notebook in GitHub. (However, the visualization for the KMeans clusters may not be visible)