Research On Segmentation Method Of Bank Individual Customers Based On Customer Value And Customer Risk

According to the WTO agreement, China’s capital market is gradually opening to the word, especial the financial market. Now, our commercial bank faces a bigger opportunity, also the challenge. Along with the increasing entrance of foreign banks, and bank-product is becoming the same, our domestic banking industry is changing from "seller-market "to "buyer-market",customer become the bank’s most important resource, the traditional "take product as the center " is urgent changed into "customer-centric". So, an advanced management idea, Customer Relationship Management, becomes the focus of research, and is already used in banks. CRM is not only a new management philosophy, also an information technology, which is supported by a large data-repository, aimed at reshaping the relationship between bank and its customers and cultivating high-quality customers. The value of customer is the core of CRM, so how to distinguish the value of customer becomes the key issue, so there needs a critical work, customer segmentation.At present, our banks have introduced CRM in some degree, but it’s not enough. What the real customer value means? How to divide individual client effectively? There is not a scientific judgment standard.In view of this, the purpose of this research is to build a customer segmentation model, which is scientific, logical and operational. It can be used to bank’s resources allocation and the customer relation marketing. Then the cooperation is becoming stable, the long-term profit of bank can be guaranteed.Thus, the general idea of this paper is as follows:First, analyze the related data; then take customer lifetime value (CLV) as the basic standard of classification, cut it into customer existing value(CEV) and customer potential value (CPV); at the same time, analyze the customer risk (CR); next, use index factor method and analytic hierarchy process (AHP) to build three evaluation systems; Finally, create a new segmentation model, and use specific data to test its validity, using the cluster analysis method of SPSS.Here, introduce the paper’s main work.First, study related theories. There has many ways to divide the bank’s individual customers through the front researches, such as profit-theory, behavior-theory and value-theory. According to their advantages and disadvantages, the paper chooses CEV and CPV to reflect the CLV; then analyzes the CR from all angles, not the credit risk only. At last, try to build a new model, using CEV, CPV and CR.Second, construct evaluation systems. This is the most important and difficult part, so it cost long time working in bank, then using index factor method and AHP to complete the work. It must consider all factors to show the basic feature of customer, including the monetary variables and non-monetary variables, subjective factors and objective factors. For factors that difficult to measure, Likert 5 level measurement is used. In brief, the three evaluation systems are built layer by layer, they reflect customer’s value and risk dynamically, thus can be used as criteria of the classification.Third, build segmentation model. It’s a three-dimensional model, which uses CEV、CPV and CR as its space coordinates, it divides customers into eight types. For each type, there has separate relationship development strategy and resource allocation policy. Especially, we can recognize the highest quality customer and mature the most potential customer from the outcome of model. In addition, it can guide the work of bank account manager.Fourth, test the model. For same reason, it takes only 20 customers as the sample. According to the above model, take the SPSS software’s K-Means method to cluster customers into eight categories. From the results, we can see that the new model is scientific and operational.So far, main contents of the research have completed and also served the purpose. In future work, our bank can plan different strategies towards each type, using the new model, not the single, unscientific standard used previously. Simultaneously, because that customer segmentation management is the principal mean and method of realizing CRM, this model may also advance the construction of bank’s customer relationship management system. Certainly, there must have some disadvantages:the process of building the evaluation systems is still quite rough; the empirical sample is much small, and so on. All of these need our further improvement.