Predictive models for customer retention in Bangladesh: enabling proactive strategies

Abstract

Ever since artificial intelligence was discovered, numerous research endeavours have concentrated on comprehending its significance inside the corporate environment [13]. These days customers are more into the quality of services provided by organisations [5]. The growing number of organizations resulted in an increase in competition and customer retention has become a major factor for businesses, as understanding its influence can aid companies to develop effective marketing strategies. The purpose of this paper is to comprehensively be able to understand the e-commerce dynamics and use relevant machine learning techniques to evaluate and use the results for the prediction of customer loyalty. The paper discusses the analysis of customer loyalty using various data mining techniques, such as decision tree, SVM, random forest, and logistic regression etc. We constructed some simple ensemble applications using the machine learning algorithms, with the dataset that we received from a Bangladesh-based e-commerce business. In the end, the abovementioned algorithms are all carried out and the result demonstrates which model is best for retaining customer loyalty.