I present the Profile Curves method as an alternative geometry-based graphical method for the problem previously tackled by the star graphs, but from a slightly different concern. The star graphs provide the marketer with strategic marketing intelligence for campaign development, as obtained from an ocular inspection and complete descriptive account of their customers in terms of variables of interest. In contrast, the profile curves provide the marketer with strategic marketing intelligence for model implementation, specifically, determining the number of reliable decile-groups. The purpose of this article is to present the method of Profile Curve for a visual display of multivariate data. I illustrate the utility of the method with an application from direct marketing.


Related Articles: Data Mining and Its Aplications