Weltkarte der Partnerhochschulen der ESB Business School

Separating the Wheat from the Chaff: Improved Variable Selection for New Customer Acquisition

Prof. Dr. Oliver Götz develops new process for campaign optimisation

Von: Lena Jauernig

No matter how good an existing customer base is – this alone is not sufficient in the long run. To stay on the road to success, a company needs new customers. But how can we identify them?

A frequently used instrument is direct marketing campaigns conducted by post. For this the company rents addresses from a specialised provider who augments addresses with descriptive characteristics such as sociodemographic details, details of place of residence or buying tendencies. In this way companies can address categories of people who might be interested in their products. The available data for the selection of addresses may contain more than 100 variables. Although the procedure is costly, the acquisition rate is often far below one percent.  

Prof. Dr. Oliver Götz from ESB Business School and his team of authors from the University of Münster and McCombs Business School at the University of Texas at Austin describe in a recent study a process with which new customer acquisition can be organised more efficiently. The study, “How to Separate the Wheat from the Chaff: Improved Variable Selection for New Customer Acquisition“, has now been published in the current issue of the Journal of Marketing, one of the world’s leading specialist journals for marketing.

What exactly is it all about? The authors evaluated data of an insurance company and of a leading address provider. The analysis yielded surprising results: On the one hand, the authors established that recipients who are most likely to respond to a campaign are less valuable to a company in the long term than customers who are less likely to respond. On the other hand, more explanatory variables for forecasting response behaviour do not necessarily lead to a more profitable campaign.   

The study demonstrates that address selection can be improved by using a so called “Bayesian variable selection model“. This innovative procedure identifies the variables that should be used for address selection, describes functional relationships between the variables and the probability of response and defines the ideal campaign size. If only a few variables suffice to separate the wheat from the chaff, companies can reduce their budgets for address selection and yet identify the most attractive prospective customers.

The results clearly show that it is worthwhile using innovative processes to acquire new customers. In this way wastage of campaigns can be reduced. Further applications of the method are the identification of customers who are liable to migrate and the selection of customers who have migrated for reacquisition measures on the basis of available information about the customers.

Link to journal


Sebastian Tillmanns, Frenkel Ter Hofstede, Manfred Krafft, and Oliver Goetz (2017): How to Separate the Wheat from the Chaff: Improved Variable Selection for New Customer Acquisition. Journal of Marketing. Vol. 81, No. 2 (March), pp. 99-113.