Radio advertising that delivers
measurable results and profitable new customers

« Previous | Main | Next »

October 30, 2006

Direct Response Radio Advertising is Science and Art

At its core, successful direct response radio advertising requires the application of principles of the scientific method. That's because in a very real way we're researchers, trying to answer the same questions to get at "what works" and "what doesn't work" for each individual campaign. The benefit of leveraging practices involved with the scientific method is that the relationship between variables can be drawn with a good degree of confidence. But that requires two elements: a) controlling all factors except the variable under study (the independent variable), and b) ensuring the use of random assignment to testing conditions.

If you're not asking the right questions, asking them in the right order, designing tests to answer those questions using as many of the principles of statistics and research design as possible, you can't expect to have a success rate that exceeds chance on a consistent basis. That's the harsh reality. If this were easy, every campaign would be a hit. The whole point of the experimental method is to reduce the uncertainty surrounding the question of whether we can conclude causality - i.e. did that change to the offer in the CTA cause the increase in response rate or was there some (or many) other factors? Did that change cause the increase in conversion rate, or was it something else?

But there's a limit to the application of scientific principles. Even in basic research (lab based), where it is theoretically possible to control all variables, research often doesn't go as planned. So imagine the challenges of "field-based" research - which is what our work most closely resembles - where there exist a high number of variables and it's nearly impossible to control or account for all of them.

Thus what's necessary to understand about direct response radio advertising is that it is impossible to conduct ourselves as true "researchers" in the way PhD researchers would identify with and approve of. That is where the "art" aspect comes in. By art I mean to refer not solely to creativity, but also to intuition that is born of years of experience and which may yield some wisdom regarding the best course to take when the road paved by scientific data and analysis ends before a clear conclusion can be drawn.

It is nonetheless possible and necessary to have our work be informed and influenced by those aspects of the scientific method because only in that manner - in a mindful and deliberate combination with our experience and intuition - does our probability of success achieve a level significantly above that of chance.

Posted by Brett Astor at 06:07 PM | Permalink