Loyal customer discriminant model using customer features on EC site

In this study, we try to measure customer loyalty attract high attention from the viewpoint of customer relationship management (CRM). From this scape there are a few studies to measure the loyalty of the customers. However, these studies have not considered the potential customer features. In this study, we focus on EC site, then we additionaly consider the behavior out of the site as well as the behavior in the site for a customer. By doing so, we caught a wide variety of customer behavior by adding the behavior data out of the site and could identify a loyal customer more appropriately. In this study, we propose a model which is able to classify a loyal customer to discover a customer who has high loyalty to the EC site. For modeling, we used the behavior data out of the EC site in addition to RFM (Recency, Frequency, Monetary) gotten from purchasing data. In addition, based on the result of the precedent study, we adopted logistic regression to analyze the purchasing history of the loyal customer identified in the discriminant model. We try to grasp the purchasing tendency of a customer having high loyalty by analyzing the purchasing data of the loyal customer. In this study, We classified loyal customer into 3 classes Customer-A, Customer-B or Customer-C. We compared the precision by the RFM model and the our RFMO model. As a result, RFMO model can more classify loyal customer and the other customer with high precision compared with RFMmodel about Customer-A and Customer-B. In addition, it was shown that purchasing tendencies is difference at each loyal customers.