A Knowledge Discovery Based Customization Services Model and Empirical Analysis

It is not feasible and practical to offer "one to one" customer service to every customer. Customer segmentation is needed urgently before drawing customization service strategy. Most stockjobbers in China classify their customers into three main groups: big customer, secondary customer, ordinary customer, according to the customers\´ possessions they declared. But this classification approach is not reasonable and effective enough, as not all wealthy customers bring high contribution value to the stockjobber while some customers who do not have a high level of investment capital account for large amounts of stockjobber\´s profits. This article shed lights on the common characteristics of the high-valued customer. After analyzing the classical CLV model, we find that customer potential contribution value to the stockjobbers largely depends on some key factors. At the basis of this, whether a new customer has high contribution value to the stockjobber in the future can be accurately predicted in their entering phase. The result of the empirical study validated this and it also reveals that it\´s possible and reasonable to differentiate high CPV customer and low CPV ones through quantitative analysis.