Conjoint experiments for direct mail response optimization

Abstract

Given the growing number of firms that use direct mail, it is important to develop methods to improve response rates. Better selection methods are useful for this improvement; determining the important characteristics of the mailing and constructing an optimal design is another way to improve the response. In this paper we focus on the latter. A framework is presented to obtain the optimal direct mail design. We propose and compare two ways of data collection, one in an experimental setting and one in a field‐experimental setting. Both approaches are illustrated with an empirical application for a charity institution, referring to the development of an optimal design of an envelope and a letter.