A Study of Carrying Out One-to-One Marketing in e-Marketplaces with Data Mining

• Main Authors: Yu-Wen Tseng, 曾郁雯
• Other Authors: Hai-Cheng Chu
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/08172006942310471574

碩士 === 東海大學 === 國際貿易學系 === 93 === Abstract Because of the popularity of Internet, more and more people go shopping on line. The cost of establishing an on-line store has decreased. But if no body visits the on-line store, it may still go bankrupt. For buyers and sellers who never meet each other, e-marketplaces are important because they can gather buyers and sellers, and provide fair information and a safe transaction environment. The competition of e-marketplace industry has become intensified, so e-marketplace makers should think of how to be outstanding among his competitors. Traditional marketing emphasizes mass production and mass advertisement. Enterprises always try to sell one product to many people. It makes marketing inefficient because not everyone needs the product. Comparatively, if enterprises understand a customer’s habit and design the personal product for him/her, they will retain old customers and attract new customers. This is one-to-one marketing. A lot of previous researches about e-marketplaces investigate how to promote customers’ loyalty to an e-marketplace or how to operate an e-marketplace, but few investigate one-to-one marketing in e-marketplace. Furthermore, by means of data mining, enterprises are able to know customers’ interest and adopt one-to-one marketing strategies. Thus, our research explores a mode of putting one-to-one marketing into practice and develops a model that can be put into use to data mining, which in turn contributes to one-to-one marketing. We can find out a buyer’s consumption inclination by his/her transaction record and recommend him/her some products that he/she is interested in. After making commodity categories and two customers’ transaction records in an e-marketplace, we apply our calculation method to analyze consumption inclinations of these consumers. The analysis suggests recommending them different products on account of their transaction records. Accordingly, our model will indeed help e-marketplace makers carry out one-to-one marketing and enable them to offer suitable products to a buyer. Keywords: e-marketplace, one-to-one marketing, e-marketing, data mining