[COVID19-BANNER position="bottom" confirmed_title="Cases" deaths_title="Deaths" recovered_title="Recovered" active_title="Active"]

Machine-Learning Techniques for Customer Retention- A Comparative Study

Machine-Learning Techniques for Customer Retention- A Comparative Study


Customer retention is the capacity of a corporation, business, product, or service to maintain its customers over a defined period. Customers who stay with a brand are more likely to return, keep buying, or refrain from switching to another service or product altogether. Businesses must establish strategies for customer retention, though, as a result of heightened competition in the telecom sector. According to studies, businesses frequently employ a variety of strategies to lower the number of clients they lose over time and enhance their experiences to keep them coming back. Due to increasing client sophistication, these strategies have become less profitable over time. Because of this, companies continue to lose clients at a rapid rate. This article compares various machine learning methods in order to provide the most effective algorithm for predicting customer attrition. Experimentation serves as the methodology. We plan to construct and train the chosen machine learning model in order to verify the effectiveness of the methods. When compared to surveys and reviews, this approach produces more accurate predictions and results. Support vector machines, the k-nearest neighbor technique, random forests, logistic regression, decision trees, and XGBoost algorithms were all experimented. Using six machine learning methods, we were able to train six models for the prediction of churn in the telecoms service sector. After training and experimenting the models with the IBM telco dataset, the Random Forest model outperformed all others in our trial, with a greater accuracy rate of 80.57%, according to the findings.

Keywords: Customer Relationship Management (CRM); Customer Retention; Machine Learning Techniques; XGBoost Algorithms; Predicting Customer Attrition

Eneh, K. M., Ituma, C., Agwu, E., & Ngene, J. N.

DOI: https://doi.org/10.5281/zenodo.7435601 | FULL PDF