Customer churn prediction using machine learning: A Malaysian telecommunication industry case study
Customer churn is a term that refers to the rate at which customers leave a business. It is affected by customer churn determinants, both directly and indirectly. Churn could be due to various factors, including switching to a competitor, cancelling their subscription due to poor customer service, o...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2023
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Customer churn is a term that refers to the rate at which customers leave a business. It is affected by customer churn determinants, both directly and indirectly. Churn could be due to various factors, including switching to a competitor, cancelling their subscription due to poor customer service, or discontinuing all contact with a brand due to insufficient touchpoints. Long-term relationships with customers are more effective than trying to attract new customers. A 95% increase follows a rise of 5% in customer satisfaction in sales. By analysing past behaviour, service providers will anticipate future revenue. This study examined which variables in the telecommunication datasets influence customer churn in Malaysia's telecommunication industry. This study aims to identify the factors behind customer churn and propose a framework for the telecommunications industry. This study applied data mining techniques to a Net Promoter Score (NPS) dataset from a Malaysian telecommunications company in September 2019 and September 2020. It analysed 7755 records with 30 fields to determine which variables were significant for the churn prediction framework. Furthermore, the correlation between variables was considered. Using Logistic Regression, Linear Discriminant Analysis, K-Nearest Neighbours Classifier, Classification and Regression Trees, Gaussian Naive Bayes, and Support Vector Machine, this study developed an integrated NPS churn prediction framework for customer churn using 23 variables. Customer churn is elevated for customers with a low NPS score. The results conclude that the Classification and Regression Tree (CART) is the most accurate for predicting churn (98%). This study is prohibited from accessing personal customer information under Malaysia's data protection policy. |
|---|