With the development of information technology, the amount of data and information we face has become more and more enormous. The technology that emerged at the historic moment is data mining, but traditional mining techniques show many deficiencies for incomplete information systems. Fully implement the group management of customer service data, correctly understand, meet and guide customer needs, pay attention to customer needs, formulate diversified service strategies, and provide customers with safe, reliable, efficient and convenient services. This is the power supply company to establish "good service", Good management, good image "three good service image inevitable requirements. Based on the experimental research method, this paper conducts corresponding research on the electricity customer data, and draws the following conclusions to describe the business characteristics of the electricity value contribution model to reflect the characteristics of different customer groups. The value contribution model is divided into 5 groups: high value contribution group, accounting for 2%; medium value contribution group, accounting for 10%; medium value high overdue risk group, accounting for 10%; ordinary group, accounting for 71%; Low-value contribution group, accounting for 3%.