Research on Customer Value Using Data Mining and RFM Analysis-Take S Department Store as an Example

• Main Authors: Ji-Cian Lin, 林繼謙
• Other Authors: Jiana-Fu Wang
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/cgi-bin/gs32/gsweb.cgi/login?o=dnclcdr&s=id=%22107NCHU5402024%22.&searchmode=basic

碩士 === 國立中興大學 === 行銷學系所 === 107 === Department stores industry in Taiwan had been growing stably and continuously for 8 years since 2009. However, it experienced its negative growth for the first time in 2017 according to the Department of Statistics (DOS). In addition to the influence from external environment, the department stores industry is also facing an increasingly saturated market and competitive environment. How could S department store develop a stable profit pattern and increase enterprise’s competitiveness in a resource-limited situation by using related techniques analyzing customer historical transaction data to find the hidden message and make the best of it to build long-term relationship with customers is currently one of the most critical issues. In this study, we firstly used data mining to filter and sort transaction records in database and executed RFM (Recency, Frequency, and Monetary) method analyzing customer value. Next, we adopted K-means clustering technique grouping the customers by their characteristics. After that, we create customer profiling according to customers’ personal information, seeking grouping feature and consumption pattern. Finally, we could offer enterprise the most appropriate marketing strategy according to the final results. According to the results of the study, the sample data was divided into five types of customer clusters. It named Disengaged shoppers, Purposeful shoppers, Behavioral loyalists, Affluent prospects and TOP customer groups. We expect this research results can provide references to both enterprise and marketing personnel for allocating resource and making decision. In the meantime, maximize the effectiveness by predict customer lifetime value (CLV) in a resource-limited situation.