International Journal of Science, Engineering and TechnologyAuthors: Mr. Parveen kumar, Dr. Deepak
Abstract: Customer churn remains a critical concern for industries such as telecommunications, banking, and subscription services, where retaining existing customers is often more cost-effective than acquiring new ones. With the increasing availability of customer-related data, data-driven approaches have become essential for understanding and predicting churn behavior. This review paper focuses on identifying significant churn predictors by analyzing demographic details, service usage patterns, and billing information. By synthesizing insights from existing research, the paper highlights how various machine learning models utilize these predictor variables to enhance the accuracy of churn prediction. Special attention is given to the role of demographic attributes such as age, gender, and location; service-related factors including plan type and usage frequency; and billing characteristics like payment history and invoice amounts. Commonly used datasets and standard evaluation metrics, such as accuracy, F1-score, and AUC-ROC, are also reviewed to provide a comprehensive understanding of model performance across studies. Furthermore, the paper discusses key limitations in current methodologies and suggests future research directions to improve real-world applicability. Overall, this review offers a consolidated perspective on effective churn predictors and provides practical guidance for developing more targeted and efficient customer retention strategies.”
