A Comprehensive Integration of RFM Analysis, Cluster Analysis, and Classification for B2B Customer Relationship Management

Customer Relationship Management (CRM) becomes more important day by day as a result of technological developments. Besides improving their internal processes for customer satisfaction, companies especially with a high number of customers seek intelligent and well-designed methods for handling their customers by analyzing their behaviors and revealing their patterns. Companies transform their repetitive processes to Robotic Process Automation (RPA) by integrating systems based on intelligence, data mining, and machine learning techniques to save time, reveal patterns in data more easily, build robust interrelationships within the functions of the company, etc. As a part of these efforts, this study proposes an integrated CRM methodology to manage both Business to Business (B2B) and Business to Customer (B2C) relations. In the integrated approach, Recency, Frequency, and Monetary (RFM) Analysis divides the customers into segments that define their current status, DBSCAN Cluster Analysis verifies RFM Analysis and reveals the anomalies in the data fur further analysis, and Decision Tree Classification Algorithm predicts company-specific custom questions to ease the work of analysts. A Case Study on the data from AN-EL A.S. (an electrical components company) verifies that the methodology produces segments of the customers and predictive responses to custom questions to enable the analysts to carry out their actions easily. According to the results of the case study, 34 % of the customers should be contacted to detail the status of the business in terms of sample approval, new purchase order, arrangement of batch size etc.