The purpose of this project is to create a model that can predict the churn of customers so that InterConnect can offer promotions and discounts to those who are predicted to leave.
We were able to achieve an AUC-ROC score of 0.844 which was above the threshold of .75. This was done with a cat boost model.
With this model, Interconnect will theoretically be able to implement marketing strategies towards customers who are likely to leave their service.
One interesting observation that we noted was that the majority of users who left did so in the first 6 months of joining. Interconnect could focus their promotions and discounts most aggressively on this segment of newer users to decrease churn.
