We present several methods for the maximization of expected profits when households are selected from a mailing list for a direct mail campaign. The response elicited from the campaign can vary over households, as is the case with fund raising or mail order selling. The decisions taken by the household are (a) whether to respond and, in the case of response, (b) the quantity of response, e.g. the sum donated or the monetary amount of the order. We jointly model both decisions and derive a number of profit maximizing selection methods. We empirically illustrate the methods using a data set from a charitable foundation. It appears that modeling both aspects of the response yields considerably higher profits relative to selection methods that are based on solely modeling the response probability.
|Number of pages||18|
|Publication status||Published - 2006|