Data defines the model by dint of genetic programming, producing the best decile table.

The Working Concepts for Building a Database Acquisition Model
 Bruce Ratner, Ph.D.

Database marketers are often tasked with finding new customers as mature markets fizzle and new markets overtake existing ones. They use models as a key component of their marketing programs towards increasing a customer database. For example, in the financial services and telecommunications industries, database marketers use acquisition models to identify individuals (prospects) who are likely to acquire credit cards or cellular services, and then develop campaigns targeted to those prospects to induce sign-up. A file of prospects is necessarily without a history of marketing information, but always includes one piece of information - the zipcode. The standard approach to obtaining knowledge about prospects is to append zipcode-level census data to their records. However, the data analyst without an advanced statistical background may not know the best method for building a model for respective prospects based on aggregated zipcode-level data. Moreover, the analyst may not be aware that implementation of a zipcode acquisition model is different from that of the usual response model.

This article discusses the working concepts for building a zipcode acquisition model by 1) reviewing the basics of weighted least-square regression analysis, 2) presenting an explicit definition of the zipcode acquisition model, 3) offering the zipcode as a link to a menu of variables that render zipcode acquisition modeling possible, 4) introducing special handling of census variables, 5) illustrating how to implement a zipocde acquisition model, and 6) providing the SAS-code program for building the model, which should be a welcomed entry in the tool kit of data analysts who frequently work on the acquisition problem.

See companion articles: 
For more information about this article, call Bruce Ratner at 516.791.3544 or 1 800 DM STAT-1; or e-mail at
Sign-up for a free GenIQ webcast: Click here.