Using Experimental Design to Improve Marketing ROI
Most marketing executives will acknowledge that their campaigns rely on far too much guesswork. The rapid pace of change and the sheer number of factors to consider make it difficult to predict how many messages will hit their targets. Some companies, however, are looking to scientific techniques and methodologies to improve their accuracy—and boost the bottom line in the process.
"Boost Your Marketing ROI with Experimental Design," published by Eric Almquist and Gordon Wyner in the October 2001 issue of The Harvard Business Review, discusses how statistical techniques long applied in other field can be adapted to predict how a marketing campaign will influence consumer behavior.
The practice of testing different forms of a marketing or advertising stimulus isn't new. Direct marketers have long used simple techniques such as split mailings to compare how customers react to different prices or promotional offers. But traditional testing becomes expensive when evaluating more than a few alternatives. Since companies now use many more marketing channels and adjust prices, promotions, and advertising messages merely by editing an electronic file, they send out an enormous stream of "stimuli."
Experimental design methodologies enable marketers to project the impact of many stimuli by testing just a few of them. Using mathematical formulas to select and test a subset of combinations of variables that represent all of the original variables, marketers can model hundreds or even thousands of stimuli accurately and efficiently.
With experimental design, companies today can collect detailed customer information and build models that predict customer response with greater speed and accuracy than ever before. As advertisers and marketers strive to break through the clutter and convert shoppers to buyers, these tools will greatly improve the odds that their efforts will pay off.