A Study to Improve the Response in Email Campaigning by Comparing Data Mining Segmentation Approaches in Aditi Technologies
الموضوعات :P. Theerthaana 1 , S. Sharad 2
1 - Department of Marketing, Anna University, Tamil Nadu, India
2 - Aditi Technologies, Marketing, Tamil Nadu, India
الکلمات المفتاحية: Data mining segmentation, RFM (Recency, Frequency, and Monetary value), CHAID, logistic regression, Email, Marketing campaigns, Response rate,
ملخص المقالة :
Email marketing is increasingly recognized as an effective Internet marketing tool. In this study, a questionnaire is constructed and distributed to a sample of 146 prospects of Aditi Technologies to find the factors associated with higher response rates. The collected data is analyzed using Factor Analysis and the 11 factors, From Line, Subject Line, Personalization of the subject line, Timings for sending mails, Frequency of mailing, Length of the Emails, Incentives to respond, Pre-existing Business Relationship, Permission based emails, Links and Image are extracted and it explains 78.363% of variance. These 11 factors is analyzed using Multiple Linear Regression and the .922 R square value indicates that 9 independent variables, Permission based emails, Length of Email, Timings, From Line, Frequency of mailing, Preexisting Business, Personalization, Incentives to respond, Subject Line contributes to higher response rate. This study also investigates marketing campaigns of Aditi Technologies using RFM, CHAID, and logistic regression segmentation methods. One-way ANOVA is used to analyze the data and it is found that there exists no difference between the three approaches. The study concludes that RFM is the most commonly used segmentation approach, however RFM may focus too much attention on transaction information (recency, frequency, and monetary value) and ignore individual difference information (e.g., values, motivations, lifestyles) that may help a firm to better market to their customers. This consideration would favor analytical techniques such as CHAID and logistic regression that can accommodate a variety of personality and individual difference information.
