Inspecting the Effective Factors on Identification and Maintenance of Key Customers Based on RFM Model and Designing a Model for Providing Services

The purpose of this research is to investigate and study the factors influencing the identification and preservation of key customers based on the RFM model and model design for the provision of services. The statistical population of the study consists of two different groups. In the first group, for determining the weight of the indicators (R F M) 18 experts from the Mellat Bank of Mazandaran province were randomly selected and for the second group in order to cluster customers based on the RFM model and using bank document data, those who were using the POS machine in 1396 were examined. The data analysis method is a fuzzy hierarchical analysis technique, entropy technique, K- Means method and DBSCAN method. According to the results, the weight of each of the RAF indexes were rated using the process of hierarchical analysis and entropy analysis and finally the weight of the indices was estimated as a combination. The weight of the indexes was M = 0.5998, F = 0.2672 and R = 0.1330. In addition, customer data clustering was conducted using K-Means and DBSCAN methods. Finally, the results showed that the K-means method is a better way to customer clustering and service delivery.