In today's competitive landscape, businesses face constant pressure to attract potential new customers. But, retaining previous customers is equally crucial, as it directly impacts revenue growth and profitability. Early detection of customer churn, also known as ‘client churn’ in some contexts, allows companies to implement proactive measures and prevent customer losses. New customers cost more in terms of customer acquisition costs compared to the retention costs, therefore customer retention is critical for business profitability. The capability of the firms to predict customer churn helps firms to identify clients who are at risk or have a high probability of switching to another provider thus enabling them implement targeted interventions. Some of these interventions include personalized discounts, improved customer service and addressing particular customer complaints not only to prevent loss of revenue but also to strengthen client relationships that will create long term loyalty. This study aims at finding the optimum machine learning approach for early prediction of client churn. Specifically, it examines various algorithms under different performance measures with an aim of offering useful information and recommendations to companies that would want to enhance their customer retention approaches. We evaluate four algorithms individually: KNN classifier, random forest classifier, logistic regression, and XGBoost classifier using established evaluation metrics. In this research paper, an algorithm is also presented for predicting the customer churn telecom industry with a greater accuracy and effectiveness.