This document summarizes the performance of a logistic regression model for predicting customer churn. The model splits customers into 10 deciles based on their predicted likelihood of churning. It shows that the model achieves lifts in predicting churn compared to random for each decile, with the highest lifts of 2.6 and 2.28 for the first and second deciles respectively. Overall, the model correctly predicts 91.9% of customers who do not churn in the first decile.