A decision support model based on customer lifetime value index and data mining: A case study

Customer Relationship Management has been an essential topic in the marketing context. Various studies have focused on CRM systems and customer data to provide insights about customers for marketing management. Customer data is collected, analyzed, reported, and utilized to model customer behavior and provide decision support for managerial decisions. The concepts "customer equity" and "customer lifetime value" provide a holistic perspective on the customer relations; thus, both are crucial for business management. In this thesis, an index has been introduced based on LRFM model that has been widely used for customer segmentation in previous research. Despite the prevalence of CLV estimation models based on RFM in prior studies, a shortage of analyses based on indexes with periodic estimation of lifetime values was noted in the literature review. It can be argued that such an index might be used in marketing management, especially for customer retention. In this thesis, a customer lifetime index was proposed based on LRFM model, and periodic LRFM combinations for customers were individually analyzed with LOF analysis for outlier detection. The CLV index and LRFM combinations were utilized as a temporal data set for various data mining techniques. A software prototype was implemented to perform LRFM analysis, outlier detection, and filter recent outliers. The model was tested on data supplied from BATIÇİM A.Ş., and noticeable outliers discovered were discussed along with LRFM components to provide managerial insights in customer retention.