Has the Age of Neuromarketing Finally Arrived?

It’s understandable why the concept of neuromarketing — the use of brain-imaging technology to better understand consumers’ behavior and preferences — gets so much attention: Predicting a product’s success from a brain scan sounds impressively futuristic, as does the idea that your neural activity is revealing something about your own likely future purchasing behavior that you didn’t know or have insight into yourself. It also seems just a touch sinister, which only adds to the fascination. In most cases, though, the hype over neuromarketing has outpaced the science. But now two exciting new papers signal that the science may be catching up.

One could be forgiven for taking a skeptical stance to media stories about neuromarketing, given the surfeit of less-than-rigorous recent articles about the subject. Two years ago, for example, Fast Company published an article under the headline “How Your Brain Can Predict Blockbusters,” ostensibly about neuromarketing research conducted by the company InnerScope. The publicity around the research suggested the company had peered inside consumers’ brains in a way that would eventually help Hollywood crack the secret code of the summer hit, but the reality wasn’t quite so exciting: The research measured movie viewers’ heart rates, breathing, and sweating, and there actually weren’t any measures of brain activity whatsoever.

This wasn’t genuine neuromarketing — it was really just wiring people up to some basic physiological measures to see how excited they were by a movie trailer. And while it’s true that other marketing research has actually employed brain recordings, the results have been mostly unimpressive. In 2010, Dan Ariely at Duke University (known for his best-selling books, including Predictably Irrational) and Emory University’s Gregory Burns published a landmark review of the state of neuromarketing. They came away unimpressed, writing, “It is not yet clear whether neuroimaging provides better data [than] other marketing methods.”

Finally, though, we seem to be approaching the genuine article, thanks to the release of a pair of research papers on neuromarketing that offer clear signs of progress. In one of the studies, brain scans helped predict whether readers of an antismoking email would bother clicking a link taking them to a quitting-advice website; in the other, they helped predict the box-office success of movies. Unlike many of their overhyped predecessors, both studies actually do show that information about brain activity can add valuable insight about future market success, above and beyond the old-fashioned method of asking consumers what they think and feel. If these results hold up and are built upon, then we’re not far from a world in which marketers can use brain scans to predict what we will buy in the future, what links we’ll click, and what movies we’ll go see.   

The more impressive — and definitely more in the public interest — of the two studies was published last month online, in Social Cognitive and Affective Neuroscience. Emily Falk at the University of Pennsylvania and her colleagues wanted to know whether functional brain imaging could be used to predict the success of an antismoking email campaign. The emails varied according to the image — either shocking (e.g., a smoker with a hole burned in their throat) or neutral (e.g., a smiling man) — that was paired with the same language: “Stop Smoking. Start Living.” Falk and her colleagues asked 47 smokers to look at each of 40 versions of the email campaign while undergoing fMRI brain scans — specifically, the researchers were curious how much the different emails triggered activity in the medial prefrontal cortex, a region associated with thinking about oneself, since earlier research has shown effective health messages trigger activity in this area. Meanwhile, 63 smokers recruited via Amazon’s Mechanical Turk survey website provided traditional market feedback: They looked at the different emails and said how each made them feel and how much it made them want to quit.

The emails were sent to 400,000 smokers by the New York State Smokers’ Quit Line, which was a partner in the study, and the real-life success of each version of the email was indicated by what proportion of its recipients clicked on a link for a website that provided help on quitting smoking. The researchers first looked at whether the popularity of the emails among the Turk workers would predict how successful they were in the real-life campaign — and, most importantly, whether also factoring in the brain-scan data would make their predictions more effective. This is an important distinction: If neuromarketing data doesn’t add to our ability to understand how a given piece of media will do “in the wild,” then it’s not particularly useful.

In this case, the brain-scan data was, in fact, useful for predicting the success of the emails containing disturbing imagery — tracking how much self-related frontal-cortex activity the emails elicited more than doubled the researchers’ ability to predict which of those messages would lead to clicks. (In stats speak, the Turk workers’ responses to the video explained 38 percent in the variability of click-through rates, while that number jumped all the way up to 65 when the brain-imaging data was factored in.) 

Summing up the results with cautious scholarly language, Falk and her team said write that “the combination of neural data with survey results may ultimately aid in more efficient prediction of the success of mass-media health campaigns.” In other words, if genuine neuromarketing hasn’t actually arrived, it’s close — and it could help us to devise health-promotion advertisements that people actually bother reading.

The other study was about movies’ box-office success. Writing in Marketing Research, Maarten Boksem at Erasmus University and Ale Smidts at the Rotterdam School of Management asked 29 participants to watch several movie trailers for films they hadn’t seen, and to then rate how much they liked each and how much they’d pay for the full-length DVDs. The researchers also recorded the participants’ brain waves using EEG (electroencephalography, which involves placing recording electrodes on the participant’s head). The key test was whether and how much the traditional data and brain-wave data would predict the real-life box-office success of the different movies.

Looking just at how much the participants said they’d be willing to pay for the DVDs, the researchers could, statistically speaking, explain just one percent of the difference between the best- and worst-performing films’ real-world box-office receipts (part of the reason this number was so low is that 29 is a small sample size compared to traditional audience research). But when the researchers also factored in measurements of the so-called gamma brain waves triggered in viewers’ brains by the trailer — waves in this range are typically seen to a greater extent when people are enjoying themselves and paying attention — this predictive power was tripled. While these figures of one and 3 percent might sound small, the researchers pointed out that “an increase of two percentage points (which represents a 200% increase in predictive power) may actually be meaningful when we consider the enormous stakes … involved in movie releases.”

Both these results are impressive in their own ways (the antismoking one, more so), but it’s important to keep in mind how early we are in the the emerging field of neuromarketing. For one thing, costs are still prohibitive in many cases: The EEG technology used in the movie-trailer research is relatively cheap and practical to use, but the evidence here was not exactly mind-blowing (part of the reason for this might be that EEG provides only crude information about where exactly activity fluctuations are taking place in the brain —  though it is very precise, more so than fMRI, at revealing when changes are taking place). The stop-smoking study, on the other hand, provided far more compelling evidence for the predictive power of brain data, but it used functional magnetic resonance imaging, which is expensive and requires people visit a research facility. That said, if brain-imaging data on a small sample could help increase the effectiveness of a large-scale health campaign, the investment could be justified and the science would surely be welcomed.

On the other hand, if this research continues apace, it’s only a matter of time before political parties and tobacco or junk-food companies start using the same techniques to sharpen their marketing messages. Like any other piece of technology, brain scans can be used for the public good, but also for less virtuous ends. In much the same way that politicians began regulating cartoon and beer advertisements directed at minors, the policy debates of tomorrow may involve what is and isn’t an appropriate use of an fMRI machine.

­ Dr. Christian Jarrett (@Psych_Writer), a Science of Us contributing writer, is editor of the British Psychological Society’s Research Digest blog. His latest book is Great Myths of the Brain.