Research On Transfer Models Of Customer Status Based On Markov Chain

The intense of the market competition and the diversification of customer’spersonal demand have made that more and more enterprises pay more attention tocustomer relationship management, but at the same time they also put forward higherrequirements to the enterprise customer relationship management. On the one hand,the intense competition for customer resources lead to the instability and volatility ofcustomer status, on the other hand, the diversification of customer’s personal demandmeans the uncertainty and variability of customer status. Different customer statusmeans different customer value. The transfer of customer status leads to the change ofcustomer value, thus in order to achieve maximizing customer value, you needcustomer status shift toward a favorable direction.This paper sorts out the theories of customer relationship management, dynamiccustomer relationship management and Markov chain to lay the foundation for futurestudy, based on the summary of the current research. Then it launches acomprehensive exposition of the customer status and status transfer. It starts from theanalyzing of customer attributes, then refines customer characteristics, and groupscustomers according to the characteristics, to determine the customer status. Then itproposes the concept of customer status transfer, analyzes and summarizes thecharacteristics and influencing factors, and classifies the status transfer from theperspective of the customer status sequences.This paper proposes the transfer model of the customer status based on Markovchain. First, it makes sure that the development of customer relationships is a longterm and has many stages, demonstrates the research applicability of the hiddenMarkov model to the dynamic CRM, then it establishes the status transfer models ofcustomer lifecycle and customer groups based on Markov chain.Finally, this paper conducts empirical research on company W. After the dataprocessing of company W, it determines the number of customer status, summarizesthe characteristics of each customer status, calculates the customer status transferprobability matrix, and analyzes the transfer rules. Then it proposes generalrecommendations to customer relationship management, and gives appropriate marketing strategies for customers in different status.