An Study Of Customer Life Value Recognition In The C2C Background Based On Rfm Model

At present, the electronic commerce has a rapid development. As a new business model,the C2C has caused a certain impact to the traditional retail. However losers are in the majority. Compared with the traditional retail Industry, online retail enterprises have less fixed assets and customers are the most valuable assets of the enterprises. In this paper, we research the customers’value to enterprises empirically based on RFM model under the environment of C2C and expect the results to help the online sellers avoid risk, allocate resources more reasonable, keep the high value customers, transform low valuable customer into high valuable customer, increase the competitive advantage.Firstly, this paper use taobao C2C shops’transaction data as sample, and reorganize the data into standard data format and analyze. We found some customers’characteristics of purchase behavior and connected it with the customers’value in order to help the subsequent research of customers’value. And then this article mainly starts from the two aspects. First of all, validate the applicability of RFM model in C2C e-business environment; secondly, discuss the mechanism of RFM under the environment of C2C. The specific research content as follows.This paper puts forward the accuracy testing method of RFM, and divides the sample into the original sample and validation samples. Verify the result of the original sample by the results of the validation sample and get the matching rate. On this basis, we summarize the method of confirming variable weight, and the variable weight will improve the model’s accuracy. At the same time,we prove that the addition model is superior to multiplication model in calculating customers’ value. We verify the applicability of RFM model under the environment of C2C by analyzing the results of matching rate. We found that there was no correlation between the matching rate and repeat purchase, which gave a lot more certainty of applicability of the RFM model in the environment of C2C. Compared with the previous researches, we found that the method has defects. It is too theoretical, and has bad practicality, through deleting customer who buys only a time as the samples and analyze comparatively with the original sample matching rate. the results showed that the matching rate of original sample is higher, and it explains that RFM model require the data of the whole customer groups. This paper studies the relationship between division of the time and the accuracy of the RFM model and presuppose that validation sample length unchanged, the original sample length changed, and then assume that the original sample length unchanged, validation sample length changed,Then observe the changes of matching rate. The conclusion is that the original sample length should not be too short or too long, and in this paper the best choice is nine months or so. Only in this way can customers’value reflect better, and advisable validation sample length is generally three months. This also showed the RMF model is not fit for long-term forecasts.The innovation of this paper is putting forward the RFM model validation methods. The study of RFM models before were many, but not inspected the precision, and this also provides new ideas for further study.