Research On Opportunity Identification Of Cross-selling Based On Data Mining

For the fierce market competition and the market saturation rate speeding up more and more, it is difficult for companies to acquire new customers, the existing customers begin to attract more and more attention. As a new introduced.marketing method, cross-selling just adapts to this changed market environment. It is a new marketing concept and idea under the guidance of the concept of market-oriented, which finds the hidden internal corporate chance by identifying exiting customers'shopping patters and identification to increase enterprise profit.Cross-selling came into being based on the customer relationship management (CRM) and database marketing which provided theoretical foundation for cross-selling. And the CRM system and other enterprises business databases provided a lot of information about customers and marketing that help enterprises to develop data mining and cross-selling, taking the customer purchase data for example, customer shopping patterns or the key factors that will influence customer next purchase even decision factors may be hidden in them.The concept of cross-selling based on the database marketing to above external environment was advanced in this study, trying to be used to solve the problems that enterprise encountering market opportunities at present.A complete process of cross-selling implementation was proposed in this study. Fist, enterprise marketing databases must be established on the basis of marketing targets, including customer databases, transaction databases and product databases; Next, cross-selling opportunities were found out from past databases; Then, cross-selling was completely implemented based on the excavated cross-selling opportunities; Last, implementation effect of cross-selling was evaluated. Among them, identifying cross-selling opportunities was focused on in this study, and the cross-selling mining algorithms from the horizontal and vertical were proposed, that were, association products were mined using association rules on the horizontal; customer purchase patterns were mined using sequential pattern analysis on the vertical, and the appropriate customer groups were searched out using the decision tree analysis according to the mined association products and sequential patterns. Moreover, the feasibility of the China's enterprise to develop cross-selling was also discussed, and the proposed cross-selling opportunity recognition algorithms were expounded by means of simulation examples.