Simultaneous Analysis of Monthly Water Clustering Shiraz Water Subscribers Based on the Value of Their Life Cycle and RFM Model

Due to the importance of identifying low water consumption subscribers, in this study, using data mining methods, segmentation and identification of water subscribers based on the life cycle value and RFM model are discussed. First, the values of RFM model indices for Shiraz water subscribers were extracted from the subscribers' database and pre-processed. After weighting these three indicators using hierarchical analysis, subscribers were segmented using self-organizing neural network; Then the lifetime value pyramid was drawn to identify key subscribers and clusters of key and valuable subscribers were identified. The results show that the water supply index is the most important and the recent payment index is the least important in determining the value of subscribers and based on this, subscribers were divided into six clusters. According to the life cycle value pyramid, the most valuable subscribers are located in cluster 3, which are mainly residential, commercial and industrial uses, are mainly residents of Zone 2, most of these subscribers have paid their bills on time and their water's amount is mostly low. By identifying valuable and key subscribers, suggestions were made to improve the service to subscribers by considering their value.