A Case Study on Machine Learning for Customer Relationship Management in Service Industry

Abstract

Data mining tools and machine learning techniques have been used in customer relationship management (CRM) for a very long time. Several papers investigated data analysis for customer retention in financial, retail, and telecommunications industries. However, there is a lack of researches on machine learning for CRM in service industry. This paper strives to understand the whole process of applying machine learning based data mining application in service industry and to examine how these novel techniques can help a business improve their customer relationship. This case use action research to document and analyze the application of machine learning based data mining in a business case in service industry. Key areas will cover decision making process from operational, managerial and strategic dimensions. The research used the data collected from a large car dealer’s IT department and its vehicle maintenance plants, containing about 2.73 million rows of data. The machine learning model used to generate the recommended customer lists was the boosted decision tree model provided by Microsoft Azure. By taking advantage of these lists, the company can increase the success rate of promoting action and decrease the time and frequency that technicians have to spend on promotion, which leads to more effective and efficient frontline operation and both higher technicians’ and customers’ satisfaction. The result of our research reveals that the recommended customer lists really helped the company better distinguish customers and achieve better CRM effectiveness through customer segmentation and customer development.