Customer Retention Based on the Number of Purchase: A Data Mining Approach

Purpose: this paper aimed at finding the relationship between the numbers of purchase and the customer’s income. The data mining tools were applied in the study to find those customers who bought more than one life insurance policy and represented the signs of good payments at the same time. Design/ methodology/ approach: in the present research the data mining tools were employed based on CRISPDM methodology. The K-means algorithm was used for classification and the prediction was based on a proposed formula in Excel worksheet. Findings: the researcher extracted some simple rules to predict customers’ clusters through selecting the customers who bought more than one policy and filtering the income- bringer customers as the companies would be able to use this prediction to change their strategies in relation to different customers. Originality/value: Utilizing data mining tools to classify different customers in life insurance and prediction based on the classification were new approaches of the study. There was not enough research and implementation in relation to the CRM and data mining in the insurance industry in Iran. Especially CRISP-DM methodology was not used extensively enough in a life insurance investigation.