Data Mining in Direct Marketing

• Main Authors: CHIN-TSUNG WU, 吳欽錝
• Other Authors: Kung Chen
• Format: Others
• Language: zh-TW
• Published: 2001
• Online Access: http://ndltd.ncl.edu.tw/handle/51147642292020929436

碩士 === 國立臺灣科技大學 === 資訊管理系 === 89 === 　　With the rapid growing of the Internet, people are using emails to communicate with each other intensively. Receiving and sending email become one of the daily activities. Since using email is more convenient, quick and cheap, enterprises also begin to use email to send campaign information instead of traditional direct mail. Marketers also use mass emails to transmit product and services information to their prospects. 　　Since direct email is low cost, marketers tend to use use email marketing to excess. They consider the more information they transmit to customer the more successful a marketing campaign. Furthermore, they send emails to all members in their hands. This is not appropriate. Email marketing should be executed in a permission-based manner. Marketers need to obtain permission from customers to send out campaign emails. Sending junk emails to customers will decrease customer loyalty and make customer reject receiving email, delete junk mail directly from their mailboxes. The mass junk emails will also decrease the click rates and reduce campaign effects. 　　This thesis proposes to use test campaign to increase the response rate of campaign emails while reducing the volume of emails sent out. The test campaign proceed as follows. First, select test members based on statistical principles. Second, send campaign emails to test members and analyze the campaign response results using data mining techniques to identify a set of potential customers. Finally, send campaign emails to those potential customers. In this way, customer won’t receive junk emails and marketing campaigns become more effective and marketer can focus limited resource on potential customer to realize highly-responded direct marketing.