An Augmented RFM Model of the Cross-Strait Consumers’ Repurchase Behavior in Online Shopping

• Main Authors: Hui-ling Chen, 陳慧玲
• Other Authors: Chin-yuan Ho
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/08524362179400335077

博士 === 國立中央大學 === 資訊管理學系 === 102 === The fast growing online shopping has turned into a battlefield for many e-commerce (EC) businesses. They must understand their customers’ purchase behavior in order to make a profit. Given the fact that the increase in customer’s retention rate can lead to higher profit and the cost of acquiring a new customer is higher than that of retention of an existing customer, the EC businesses can understand their customers’ behavior and assess customers’ value in order to initiate target marketing or precision marketing by capturing the probability of revisiting the same seller by a customer and repurchase at the same e-marketplace. Taking China’s largest EC platform—Taobao, and Taiwan’s top two platforms—Yahoo Taiwan Auction and Ruten Taiwan Auction as our research targets, and focusing on the most popular trading categories—women’s apparel, we conduct a comparative analysis on the cross-strait EC consumers’ repurchase behavior. The purpose of this research is to establish a RFM-based prediction model of consumers’ seller repurchase and platform repurchase by analyzing the actual transaction data of women’s apparel and to compare the cross-strait EC consumers’ repurchase behavior. The repurchase behavior prediction model consists of five predictors, including the recency, the freguency, the total amount, the average amount, and the consumer’s last rating. The research findings show that in terms of repurchase rate, Yahoo! is the highest, followed by Ruten, and Taobao is the lowest. Interestingly, the consumer’s seller switching rate in descending order is also Yahoo!, Ruten, and Taobao, which indicates the consumers at Yahoo! exhibit multi-loyalty behavior with both high repurchase rate and high seller switching rate. The Logistic regression shows that all the predictors in the seller repurchase and the platform repurchase prediction model of Yahoo!, Ruten, and Taobao are statistically significant. We also use cluster analysis to identify the characteristics of the most valuable customers at the three different platforms. All of our findings are based on actual transaction data of online shopping web sites, the repurchase behavior of online consumers and its prediction model can be used by EC businesses and platform businesses for consumer relationship management and merchandise sales and marketing.