What Don Draper could do with an fMRI…

Humans are terribly inaccurate at predicting their own behavior. Yet, marketing and advertising industries have historically relied upon individuals’ predictions about whether they would use select products, when, and why via focus groups and marketing surveys. 

Luckily, advances in psychological science, in this case with the use of fMRI, we have a window into how the brain responds to the world that is, in some ways, independent from our subjective report. Thus, inquiring minds were eager to test whether brain activity might be a better predictor of our behavior than our reports. For example, in 2011, Dr. Emily Falk and her colleagues published a study showing that neural activation in the medial prefrontal cortex (MPFC), an area involved in self-referential thinking, while watching advertisements for smoking cessation was a better predictor of quitting smoking than self-reports. This is pretty fascinating on the surface, but also very intuitive. People lie. If anyone appreciates that fact, it’s psychologists. We have dedicated years of research to developing ways of patterning the way people lie in order to more accurately measure psychological constructs.

There are a lot of reasons people who want to quit smoking might have inaccurate self-reports on whether they will quit smoking. First and foremost, doctors want their patients to quit smoking, so smokers tell their doctors what they want to hear. However, Dr. Falk and her colleagues conducted a follow-up study where they were interested in whether neural activation to smoking advertisements would predict the behavior of others. In other words, they asked the question:

Can neural responses to advertisements in a few people predict how effective they will be for others? 

To answer this question, the research team recruited 30 individuals (15 males) who were about 44 years old. Each of these individuals were “heavy smokers” with “strong intention to quit” smoking. Each participant then watched 16 smoking cessation advertisements during an fMRI session, where images of activation in their brain were taken. These advertisements were part of 3 different ad campaigns (campaign A, B, C) sponsored by the National Cancer Institute’s Smoking Quitline (1-800-QUIT-NOW). These were real advertisements that were aired locally in a different part of the country, and none of the participants had seen them before. After watching all of the advertisements in the fMRI, participants rated each advertisement’s effectiveness, including whether it motivated them to quit and how persuasive it was. Then, the research team collected data from the Smoking Quitline on how effective each advertisement actually was. To do this, they compared the number of calls they received during the month before each advertisement aired with the number of calls during the month after that advertisement aired.

Based upon the participants’ ratings of advertisement effectiveness, the most effective advertisement campaign was campaign B, followed closely by campaign A, and then campaign C by a larger margin. In contrast, the MPFC demonstrated the greatest activation when watching campaign C, followed closely by campaign B, and then campaign A. So, the brain and self-report are telling us something different, and Dr. Falk’s previous paper suggests that the neural activation in this region would more likely predict who will actually quit smoking. So, which was correct?

Based upon the change in calls to the Quitline before and after each advertisement aired, campaign C was by far the most effective. Campaign C resulted in a 32-fold increase in call volume to the Smoking Quitline, while campaign B resulted in a 12-fold increase, and campaign A only resulted in a 3-fold increase. This effectiveness ranking between campaigns (C > B > A) matches that predicted by neural activation in the MPFC.

Many companies spend millions of dollars creating, filming, and airing advertisements, especially for big events such as the Superbowl (This was my favorite Superbowl XLIX commercial, by the way: #likeagirl). These millions of dollars are intended to be an investment in the even larger revenue generated by the advertisement. Based on this article, the time may be quickly approaching when experimental approaches using fMRI may extend into the private sector, helping advertising agencies determine which campaigns will give them the most bang for their buck.

Obviously, this study was very specific to a target audience where self-reports are perhaps more susceptible to bias than consumer products like which dish soap you buy. It remains to be seen whether neural activity to an advertisement in a group of random individuals in the community, including smokers who have no intention to quit and non-smokers, would return the same results. Given that the brain region with the predictive value (MPFC) is associated with self-referential thinking, I would expect that this finding is limited only to sub-groups of individuals for whom the ad is relevant. However, that’s not too different from marketing focus groups; they use housewives for focus groups on laundry detergent and athletes for focus groups on high-performance running shoes. To truly compare, they would need to conduct this study again comparing the predictions of focus groups on ad effectiveness with the predictions based on neural activity.

Even so, this team of creative researchers still have strong evidence that what a person’s brain is doing in Los Angeles when watching a new advertisement will predict whether a person in Louisiana will call a hotline after seeing the same advertisement. And that, my friends, is very cool.

Falk, E. B., Berkman, E. T., & Lieberman, M. D. (2012). From neural responses to population behavior neural focus group predicts population-level media effects. Psychological science, 23(5), 439-445.