With the progressing of marketing and growing of information technology,all industries are so competitive.Whether to maintain a good relationship with the customers will directly affect the development of a company.Customer Lifetime Value is the core of Customer Relationship Management.Both academy and enterprises pay attention to the modeling and applying of CLV.Sometimes extrinsic or intrinsic factors affect the experience negatively in service,which will lower the expectations of the customer.The customer who experiences a poor service will complain to the company for the compensation.There will be two results after claim,the customer will be comforted when the feedback meets his expectations or lose when his expects not reached.We want to study the customer behaviors with complaints for its innovative points.In this paper,we focus on building model based the actual transaction data and combining the customer base analysis model and the monetary value model for CLV calculation.Then,the group of customer behavior is founded by means of customer value segmentation.Firstly,Logit model is built for discovering the complaints have a significant impact on customer behaviors.Secondly,BG/NBD model is improved as a customer base analysis model in this paper for its better predicting effect.Thirdly,Gamma-Gamma model is built for the final CLV model.Finally,we group the customer based Customer Value Matrix Theory.Our research finds that complaints actually affect the customer behaviors and the model improved real fit.Clarifying the classification of clients can provide advice for the management decision in the enterprise. |