Predictive Analysis of Customer Retention Using the Random Forest Algorithm

Abstract

Retaining customers is becoming a measurement focus in an industry with increasing competition. The concept of customer retention has become a research study in the sales industry, because it is difficult to retain customers and easily switch to other brands. Customer repurchase decisions in the business world of sales are very competitive. Customer satisfaction is directly proportional to the retention rate, if the customer is not satisfied then the automatic retention rate will be low. If the company is not able to meet customer expectations, it will have a serious impact on the company, namely moving customers to other services. Service factors, price, profit value, satisfaction and trust affect customer retention. One of the factors that influence consumers to become customer retention is service quality. A predictive customer retention plan is needed with data mining using the random forest algorithm. The random forest algorithm is a method that generates a number of trees from sample data, where the creation of one tree during training does not depend on the previous tree, the decision is based on the most voting. The voting results from several decision trees that are formed are the boundaries that are used as class determination in the classification process and the most votes are the winners and determine the classification class. This study aims to determine and analyze customer loyalty, customer trust and customer satisfaction. So that it can make it easier to monitor customers at the company. The results can be seen with the percentage of about 81.12% customer retention and about 18.87% customer churn. The result of feature evaluation shows that customer_activity has the highest influence on customer retention, followed by subtotal and qty.

Analysis of Factors Affecting Purchase of Self-Defense Tools among Women: A Machine Learning Ensemble Approach

“…With that, it was concluded that the RFC algorithm is the best model for predicting consumer behavior, having an accuracy of 94.3% with 19 decision trees in total. Furthermore, Suhanda et al [52] showed that the results and predictions of the RFC algorithm with a 96% accuracy rate. This concluded that customer activity has the strongest influence, causing customer retention implying that consumers' behavior influenced their purchase intention.…”