Research On BG/NBD Prediction Model And Its Application Of The Customer Purchase Behavior

Along the development of the technology, the social production is improved quickly and the company's pattern changed. In many vocations, it is difficult for the companies to win in competition just depending on the quality, price and the difference."Focusing on the product"is replaced by"Focusing on the customer". The companies have waked up to the fact that the customer is important to the market competition and their development. The core advantage is to keep good relationship with the customers. Now, the customer relation management has become one aspect of the company management. The customer behavior prediction is the basis of the customer relation management. It involves the application of the prediction model.Aiming at the BG/NBD prediction model, this paper tells us the background, the importance and the actuality of the study first. Then, we analyze the influence factors of the customer behavior prediction and their influence degree. Then, we analyze the existing prediction modes, Dwyer method, customer event method, simulating method, SMC model and BG/NBD model. This paper, we just introduce the front models simply, emphasizing on the analysis of SMC model and BG/NBD model. The data that we use in this paper is from one medical treatment company. We divided the data twice, prediction data and inspection data. According as the prediction data, we use the model to predict their succulent behavior and then compare with the actual transaction behavior. We use the degree of the activity and the expecting transaction numbers in comparing them. On basis of comparing them, we conclude that the prediction ability of these two models is similar. In calculation, the BG/NBD model is more convenient. So, the BG/NBD model is more suitable for actual implication.