Dynamic Customer Relationship Management, Hidden Markov Modeling And Empirical

The effective customer relationship management is the main route that the company keeps the competition advantage. The main jobs of customer relationship management which companies considered include four parts, such as, identify, subdivision, promote, maintain customer. The contents can be simplified to two problem: who will buy our products? How many he/she will buy? To solve the problem, many scholars focus on the customer'behavior for calculating the customer's value.The early model only calculate the behavior data which customers has, and predict the purchasing probability next term. Base on this model, some scholars consider more factors to calculate the purchasing probability, such as, sales promotion and customer satisfaction. In practice, however, present studies didn't match the needs which the companies face. The company not only want to know whether cheapen next term, but also want to know how may times he needs to carry out sales promotions. So to solve the problem, we need the dynamic customer relationship management.This dissertation addresses the issue of modeling and understanding the dynamics of customer relationships. The proposed model facilitaties using typical transaction data to evaluate the effectiveness of relationship marketing activities as well as the impact of past buying behavior on the dynamics of customer relationships and the subsequent buying behavior. My approach to modeling relationship dynamics is structurally different from the models in the existing literature.The model has two following characteristic:Firstly, the exiting models only use trade to scale the relationship between customer and company. In this model I used more factors estimate the relationship, has not the factor numbers limit. Secondly, we can use the relationship states as the standards to subdivision customers dynamically.In addition, I calibrate the proposed model using customer's buying data provided by an B2C company. This empirical application demonstrates the value of the proposed model in understanding the dynamics of customer-company relationships and predicting buying behavior.