Clustering and Prediction Model of Customer Lifetime Value (Case Studies: IRAN National Center for Numbering Goods and Services)

One of the important topics in this field is Customer Lifetime value(CLV) that means how much profit a customer generates in his lifetime for a company. The main goal of this research is presenting a model for clustering and predicting customer lifetime value and customer evaluation in IRAN National Center for Numbering Goods and Services. In this research, 74,385 records of members at a specified time interval were used (from 2011 to 2017). Members are classified by CRISP methodology, resulting in the presentation of a model for predicting them. At first, members are classified into 7 clusters by RFM and K-Means. Next, each cluster is rated by CLV. Next, hidden patterns are discovered inside the data and various segments of members are then predicted through classification algorithms. Finally, the algorithms are evaluated and the final report is prepared. The results of this study exhibit the member’s behavior in each cluster on the organization’s services and membership or extended subscription and also future customer behavior are unveiled. This research helps the managers to come up with marketing strategies to keep loyal members and attract or remove inactive members by analyzing the clusters.