Font Size: a A A

The Empirical Research Of Potential High Value Customer Mining Based On Customer Segmentation

Posted on:2018-04-03Degree:MasterType:Thesis
Country:ChinaCandidate:Y Y BaiFull Text:PDF
GTID:2359330518468750Subject:Master of Statistics in Applied Statistics
Abstract/Summary:PDF Full Text Request
With the rapid development of our economy,"beauty economy" has gradually become a hot topic of consumption.The number of people using cosmetics is increasing,more and more demand for cosmetics,and this trend prompted the rapid development of luxury cosmetics in China.All international luxury cosmetics company competitive,so strengthen relations with customers,effective mining and manage customer resource is the key to gain market competitive advantage,but in terms of the cosmetics company(hereinafter referred to as the GL company),has a lot of customers,and high value customer recognition is very high for their products,the price is not very sensitive,do not need to do too much maintenance,so the company policy makers are concerned about the future may become the high value of customers,known as a high value of potential customers,and to make personalized marketing strategy to them.But how to excavate potential high value customer,and analyze the customer's preference? This is the most concerned about GL company policymakers.This paper draws on the customer life cycle theory and the thought of the long tail theory to mining potential high value customers in the GL company customer's tail,and making personalized sales strategy.From the customer life cycle theory,shorten the customer's review and formation,as soon as possible into the plateau,to increase customer's contribution to the enterprise value in the whole life.This paper is to use the GL company customer transaction data,using the RFM model of extension model RFMD scale model to determine the GL company high value customers and low value customer,and according to the customer to purchase other information,using the decision tree and support vector machine(SVM)algorithm,learning high value customers and low value clients buy characteristics,to mining the potential high value customer,and connecting with the customer purchase behavior of activities and products for customer segmentation,Find out all potential high value customers like the products and activities,you can send the information for product or activity of interest to customers,to provide decision basis for marketing decision makers,they can make them more rational decisions that potential high value customers personalized sales strategy.
Keywords/Search Tags:Potential high value customer, RFM Model, K-Means algorithm, Support vector machine, Customer Segmentation
PDF Full Text Request
Related items