Abstract

Nowadays, customers are the essential elements of marketing for business operation. It is a critical and unignorable task in exploring valuable customers for companies and estimating customer values. According to the definition of Customer Life Value (denoted as CLV), a suitable model was found in this study and customers’ present values were estimated by given data from insurance company. Two data mining technologies (Rough Sets Theory and Decision Tree) were introduced and applied to find the rules and factors which might have influence on customers’ values. The comparing results of two technologies revealed that the influential and important factors for both technologies were similar but not for the minor factors. Both technologies performed well in efficiency of analysis but were different in interpreting results.  It suggested that the rules generated from both technologies could serve as the auxiliary factors in practical marketing strategies.

Key words: Customer lifetime value, insurance industry, rough sets theory, decision tree.