The Research Into The Car-hailing Customer's Value Based On Improved RFM Model And Clustering Algorithm

With the continuous development of the car-hailing industry,the market has gradually stabilized and the relevant enterprises have gradually entered a mature period.In this context,the only way for the car-hailing platforms to gain its profit is to segment customers,only by doing that they can develop the marketing and management strategy in accordance with the characteristics of customers to ensure the optimal resource allocation efficiency.It is an effective way to segment customers from the perspective of customer value.Therefore,clearing the way to measure customer value of cae-hailing,building a customer segmentation model based on customer value and realated CRM strategy is the primary problem to be solved.First of all,starting with the present situation and trend of the development of the car-hailing industry,this paper establishes a RFMS model which balances the perceived value of enterprises and the perceived value of customers.Secondly,the R,F and M indexes of customer value are respectively clustered and the clustering rules are extracted to subdivide the car-hailing customers.Taking into account the business focus of current development of the car-hailing platform,the customers are divided into high value,development,maintain,potential and low value of these five types of customers,a statistical analysis of their attributes and behavior is conducted to tell the characteristics of various types of customer segmentation.Then,we combine the customer's RFM division and S division,based on the level of perceived value of both the platform and the customer,the customer will be divided into the core customers,customary customers,support customers,general customers and risk customers,and the devided customer's characteristics of a qualitative description will be given.Thirdly,we use the AHP to determine the index weight of the customer value score,taking into account the influence of the difference of the index value on the index weight,the index weight is weighted in stages to ensure the scientific and reasonable of the result.Finally,we use the index weights to calculate the value of a single customer and also use the method of 28 principles to cross-validate the result of customer RFMS segment and the customer value.The correctness and rationality of the result of division and the result of value calculation are illustrated.The innovation of this paper lies in improving the traditional RFM model taking into account the customer perceived value in the area of car-hailing,and carrying out phased empowerment of the indicators for calculating customer value.This model has a positive practical significance on strengthening the customer value analysis of the car-haling platform,improving management and the implementation of efficient CRM strategy.