An intelligent Prediction of Customer Churn in Telecom Industry Using Business Analytics Classification Models

There are two common challenges faced by any internet service provider: the first being attracting new customers to grow their clientele, the second being retaining the existing customers and prevent them from leaving the company. It is established that client retention is six times cheaper than the cost of acquiring a new one. In order to achieve this, companies are increasingly investing their time and channeling their efforts to keep their clients on board. Understanding client behavior is key to identifying which of them have a high potential of leaving the company, thereby helping businesses develop appropriate retention strategies in response. This client attrition is commonly known as customer churn. Our proposed work focuses on churn prediction on Orange Company Client Dataset using four business data analytics models: Logistic Regression, KNN, Random Forest and Decision Tree. Based on the results, Random Forest boasts with the highest accuracy and F-1 score emerging as the clear winner.