This study examines predictive analytics for customer churn using a simulated dataset like the Telco Customer Churn dataset with information about contract type, service usage, fee per month, tenure, and support contacts. Machine learning models like logistic regression, random forest, and gradient boosting are employed to ascertain the most influential determinants of customer retention. The analysis showed that the determinants that best predicted customer churn are contract type, tenure, monthly fee, and support contacts. The best model was the one based on the use of gradient boosting that achieved the highest rate of 85% with an F1 measure of 0.80, which signifies higher predictive ability than the random forest model, which achieved the highest rate of 82%, and the logistic regression model that achieved the highest rate of 78%. The study shows that optimizing responsiveness to customer support and using longer-duration contracts can effectively suppress customer churn rates. The study provides practical recommendations to service companies, particularly those using the Salesforce Service Cloud, to enhance customer retention through proactive targeting of churning customers.