Examining Two-stage Modeling for Customer Segmentation Using Intelligent Clustering Techniques

Recognizing the value of customers is a key factor in the success of different stores, and it has become more prominent today. Chain stores are in contact with different customer groups and considering their limited resources, they need to rank customers based on their values to be able to allocate an appropriate portion of their marketing resources to more valuable customers to earn more. Therefore, we use data mining techniques to sort customers. Much research is done on this issue. In many studies, the RFM model is used to classify customers. This model consists of three indicators of recency, frequency, and monetary value to analyze customers’ purchasing behavior and can determine customers’ behavioral value. In this paper, a comprehensive method using three segmentation models based on RFM, SOM, SODA, VFT, and k-means models is provided, and for customer identification, transactional and demographic data has been investigated. Proposed models were implemented in chain stores and 347 customers were studied. For transaction data, transactions recorded in the store’s information center were used and demographic data were also asked from every customer on the phone. The customers were classified using each of the three models, and in the end, these models were evaluated and compared by the Davies-Bouldin index (DBI) and the sum of squares error (SSE). Based on DBI, the first model showed better performance in this case study, but based on SSE the second model performed better, and this difference is due to the nature of these two indices.