Author: Dietrich Eherler, Thomas Lehmann
Lehrstuhl für Wirtschafts- und Sozialstatistik
The identification of responder profiles is one of the most important issues in marketing research. In this
paper two approaches are applied to deal with this problem, illustrated by a case study from the automobile sector. The objective of the direct mailing considered here was the acquisition of new customers. Therefore persons similar to the existing customers of the company had been contacted with a brochure of the products. We will investigate if the achieved response rate could have been increased by the application of the suggested approaches. Starting point is information about this former mailing to potential customers. This information consists of the response behaviour of the addressed individuals as well as demographic, geographic and product specific data.
One widely used method for this kind of problem is CHAID analysis. We compare the results of this approach with those obtained by dependency analysis, a method which is still rarely applied to this issue. It turns out that CHAID and dependency analysis are appropriate tools to classify responders and non responders by their attributes and thus to improve the performance of direct mailing activities.
Full Text: lehmann