Customer Risk Research For CLV Modeling In B2B Context

In the last decades, CLV (Customer Lifetime Value) has got increasely attention in both academic and industry cicles. These reseach are basicly focus on individual customers, whose risk for company are not evidence. However, in B2B (Business to Business) situation, customers are enterprise customers, which means smaller number of customers and larger transaction with one single customer. Compare to individual customer, a enterprise customer’s risk may create more losses for a company. Therefore, this paper focus on how cusmoter risk adjusts CLV in B2B situation, which not only is a reseach gap in academic area, but also has pratical meanings on enterprise customer mangement in industry cicle. The main contributions of this paper include the following three points.(1) Selected several kinds of risk that affect CLV and formulate metrics to compute them. In this paper, the author has developed selections rules, and them picked up four risk types, include wave risk, fall risk, churn risk and credit risk to adjust CLV. Each kind of risk has been given the calculate metric.(2) With a improved Bayes Network, this paper has reseached the relationships of there risks and merged them into one single risk score. To get the final risk score, the author has abandon the traditional Data Mining methodology and improved existing Bayes Network methodology. The Customer Risk Bayes Network has been instrucured to calculate the correlation of different risks and the final risk score of each customer.(3) Adjusted the CLV model with risk by learning the idea from Beta risk. The existing CLV models are coming from Capital Assets Pricing Model(CAPM). Since CAPM lies in Financial area, this paper adjust CLV with risk in a financial way. Transformed the risk score-which is the output of Bayes Network-into Beta risk score, at last, adjusted the discount rate in the basic CLV model with risk.