Neuromarketing: Prove Thyself & Protect Consumers


by Ann Parson

December 19, 2011

BRIEFING PAPER


Ann Whitman
(212) 223-4040
awhitman@dana.org

Last year, for the first time in decades, shoppers encountered a new Campbell Soup label on supermarket shelves. The metallic spoon was gone, and the new label featured a bigger, more contemporary bowl filled with steaming soup.[1] This marketing gamble was largely due to research by Innerscope Research Inc., a Boston-based neuromarketing research company, on consumers’ responses to the old label. Eye tracking and pupil dilation tests of forty people, along with biometric measurements of heart and respiratory rates, sweat levels, and changes in body posture, indicated this warmer image might sell more soup.

Neuromarketing, the practice of using neuroscience to try to determine a person’s unconscious biological reactions to a product, “is here to stay,” said Judy Illes, Ph.D., Canada research chair in neuroethics at the University of British Columbia and a Dana Alliance for Brain Initiatives member. Reportedly, a growing number of advertisers and media firms are hiring neuromarketers to help with their campaigns. At the same time, the advertising industry itself wants neuromarketing firms to exhibit their scientific methods more openly and act responsibly with their claims.

The problem is that researchers performing neuromarketing studies are mostly in private companies and under no obligation to publish their work. “In many cases, we have contractual obligations to keep things private for our clients,” said Carl Marci, M.D., Innerscope’s CEO and chief science officer, as well as the first director of social neuroscience at Massachusetts General Hospital’s Psychotherapy Research Group. (Marci, it should be said, bends over backwards to explain Innerscope’s methods, which include frequent presentations and papers in the International Journal of Advertising and Journal of Advertising Research.)

Illes, for one, thinks that both public and private sector researchers have a moral obligation to advance sound science and disseminate new knowledge. “I would argue we need a code of ethics for anything we do that involves brain technology,” she said. That includes full disclosure of neuromarketers’ methods and accurate representations of their technologies to the public.

Gathering the Data
Although tests aimed at peering into a person’s unconscious have been around since Sigmund Freud, what brought neuromarketing research alive, said Marci, is the ability to gather immense amounts of data from an audience, aggregate and process that data rapidly by computer, and see patterns emerge that give a sense of whether or not the audience is emotionally engaged at any given moment.

For the Campbell Soup assignment, technicians fitted forty participants with Innerscope’s biometric belt, which simultaneously collected four channels of data: heart rate, sweat, respiration, and motion. Researchers collected upwards of half-a-billion data points, then ran them through algorithms that provided insights into activity within the participants’ subcortex and limbic systems, regions sometimes referred to as the brain’s emotion centers.

“Today’s brute force approach of computationally mining brain-related data is analogous to studying the human genome,” said Gregory Berns, M.D., Ph.D., a neuroeconomist at Emory University and co-author of a recent paper in Nature Reviews Neuroscience, “Neuromarketing: The Hope and Hype of Neuroimaging in Business.”[2] “We do not understand what genes do in most cases, or how they interact with each other. However, you can mine large sets of gene data and find things that correlate with disease or behavior, and the same trend is happening in neuroimaging. Through data mining, we can find patterns of brain activity that predict behavior or cognition without necessarily understanding what the patterns mean.”

Before he took up neuromarketing, Marci spent ten years applying biometrics to the study of empathy, and came upon interesting instances of physiological concordance between two individuals facing each other during a clinical encounter or other social interaction: signals tied to their biometrics, or physical and behavioral characteristics, would line up. Researchers thought these moments occurred because of the mirror neuron system, [3] upon which Marci eventually based the algorithms that Innerscope now uses to gauge an audience’s synchrony and intensity of response when viewing a product.

“If two people can generate moments of shared meaning, imagine what happens when an entire audience watches an engaging story,” like a television advertisement, said Marci. “Their mirror neuron systems will take everyone in the audience on a common journey, and the better that journey is, the more we can tell what information in that advertisement matters.”

Marci refutes the notion of a “buy button” in a person. That would imply there’s a certain region in the brain that neuromarketers could activate through product design that would make people rush out and buy a product. As it stands, even if neuromarketers can enhance a product’s appeal so that, say, it activates the striatum—a region many studies have correlated with pleasure—that might only negligibly raise the odds of someone buying something. For neuromarketers and their clients, it is still not clear if that is useful information.

Does the Science Support the Claims?
How real is the science behind the practice? And does it work? Ever since an Atlanta advertising firm first used the term “neuromarketing” in 2002, qualified investigators outside the field often say they can’t tell.[4] 

The uncertainty relates to whether neuromarketers can accurately translate their data for the sake of answering real-life questions like, “If an fMRI reveals that a group of forty people experience far greater activity in the brain’s ventral striatum when shown a photograph of President Obama than when shown a photograph of Mitt Romney, does that mean they will vote for Obama?” (Indeed neuromarketers’ claims extend to discerning voters’ preferences. Someday, conceivably, their insights might even guide the designing of a candidate’s appearance and message content.[5])

“It isn’t safe to assume that because a brain area is active, that means that the function commonly associated with that region is engaged,” said Carl Fisher, M.D., a third-year resident in psychiatry at Columbia University and lead author of a 2010 Harvard Review of Psychiatry paper on neuromarketing.  In the case of the striatum, said Fisher, “that region has also been associated with aversive stimuli, like pain, so who’s to say the participants aren’t Tea Party supporters who are anticipating that Obama will do something they find distasteful, like raising taxes? This is true of other brain regions. Any inferences linking specific regions to specific cognitive functions are speculative, because so many different patterns of brain activation have been associated with so many different cognitive functions.”   

For this very reason, back in 2007, when researchers from the University of California, Los Angeles, and others published an op-ed in The New York Times, in which they tried to infer the political leanings of twenty voters by using fMRI to watch their brain activity, many neuroscientists immediately dismissed the study as seriously flawed.[6] Photos of Mitt Romney had elicited “a significant amount of activity in the amygdala, indicating voter anxiety,” the researchers had noted. Yet, as prominent neuroscientist and Dana Alliance member, Martha Farah, Ph.D., University of Pennsylvania, blogged that same day: “The Romney amygdala activation might indicate anxiety, or any of a number of other feelings that are associated with the amygdala– anger, happiness, even sexual excitement.”[7] Farah described images of brain activation as “ambiguous” and open to interpretation as “tea leaves in the bottom of a cup.”

For their 2010 neuromarketing paper, Fisher and his colleagues reviewed the websites of sixteen neuromarketing firms and reported the following: Thirteen companies “described their methodology, but these descriptions were often insufficient to determine what was being done.”[8] Altogether, the Columbia team uncovered “a paucity” of peer-reviewed reports on the company sites, most of them dated. Eleven websites referenced none at all, and only one company “provided citations for its specific claims.” Nine of the companies, however, listed staff members with advanced science degrees.

Although ten companies promised “the truth” in discerning customers’ preferences, Fisher and his colleagues came away skeptical. “Companies may be making premature claims about the power of neuroscience to predict consumer behavior,” they concluded. Fisher added, “On the basis of what was available to us, we didn’t see a lot of evidence that neuromarketing works or is effective, which gets back to the original problem”—that private companies have no incentive to publish their science. “I don’t see their papers answering any specific questions about how to market Fritos over Cheetos.”

Still, many of those who investigate neuromarketing are cautious about categorically ruling out its potential, given that the field is new and the science as complicated as the brain itself. “It’s hard to be black or white. If someone takes a hard line against neuromarketing, I would say, ‘Well, what’s that based on?’ because there’s not a whole lot of data in the first place,” said Berns.  

The Advertising Field Weighs In
In response to its members asking questions about neuromarketing, in 2010 the Advertising Research Foundation initiated the ARF NeuroStandards Collaboration, with the goal of evaluating neuromarketing services and developing guidelines. Independent academic reviewers, including neuroscientists, reviewed eight neuromarketing vendors and their methods. A large circle of consumer brands, media companies, and advertising agencies sponsored the nine-month project, at least a few of which—like those working with Campbell Soup Company and Warner Bros.—already make use of neuromarketing.

The resulting white paper doesn’t shy away from criticism: The potential for tapping a person’s unconscious is real, it says, particularly in light of recent advances in neuroscience and technology, but “the complexity of the science underlying these methods makes it difficult to assess their validity.”[9] The project’s reviewers critiqued what they saw as neuromarketers’ too-frequent exaggeration of what their tests could deliver and felt that “documentation of methods, research protocols, and clarity about what was done are essential,” given the complexities involved.[10]

Annie Lang, Ph.D., an Indiana University psychophysiologist and a senior reviewer for the ARF project, said that while some claims were less than convincing, “the claims of a couple of companies were reasonable. Their measures seemed valid; they did good statistical testing, delivered good inferences, and they were appropriately cautious with their conclusions.”

Identifying the parts of advertisements that elicit cognitive and emotional states “can be useful,” said Lang. But, she added, “My career has been spent scientifically assessing whether we can make reliable inferences about what people are thinking and feeling while interacting with media, based on real-time measures, and it’s a pretty hard thing to do. These days, anyone can collect the data. How they interpret those data is another story, and, in many cases, we can’t assess their claims, because they’re based on in-house proprietary data.”

What’s Next?
Before neuromarketers make further inroads into the marketplace, Illes recommends that the industry adopt a five-star code of ethics: 1) protection of research subjects, 2) protection of vulnerable populations from marketing exploitation, 3) full disclosure of goals, risks, and benefits, 4) accurate media and marketing representation, and 5) internal and external scientific validity of neuromarketed products.[11]  

Illes and other industry watchers look to the future when regulations might go a long way toward strengthening the science behind neuromarketing, its validation, and transparency. In the meantime, ARF recommends that new and old marketing research practices be used together in complementary fashion. It’s unlikely that advertisers who can afford the newer methods will stick only with traditional practices, especially given a widely held observation: The vast majority of emotional processing that consumers’ choices are based on, say neuroscientists, takes place at the unconscious level, and only newer methods, not traditional ones, are capable of plumbing those depths.

Published December 2011


[1] The Wall Street Journal, February 17, 2010; Advertising Age, July 24, 2011.
[2] “Neuromarketing: The Hope and Hype of Neuroimaging in Business,” by Dan Ariely and Gregory S. Berns, Nat Rev Neurosci, April 2010, 284-292.
[3] The mirror neuron system is defined as a network of cortical neurons that is critical to a person’s ability to perform an action as well as recognize the consequences of that same action when performed by someone else. However, conclusive evidence of its workings is still lacking.
[4] The word neuromarketing first appeared in a press release in June 2002. Source: “Defining Neuromarketing: Practices and Professional Challenges,” by Carl E. Fisher, Lisa Chin, and Robert Klitzman, Harv Rev Psychiatry, July-August 2010, 230-237.
[5] Dan Ariely et al.
[6] “This is Your Brain on Politics,” The New York Times, November 11, 2007.
[7] “This is Your Brain on Politics?” by Martha Farah, for the Neuroethics & Law Blog http://kolber.typepad.com/ethics_law_blog/2007/11/this-is-your-br.html
[8] “Defining Neuromarketing: Practices and Professional Challenges,” by Carl E. Fisher, Lisa Chin, and Robert Klitzman, Harv Rev Psychiatry, July-August 2010, 230-237.
[9] NeuroStandards Project White Paper, AFR NeuroStandards Collaboration Project 1.0, p. 7.
[10] NeuroStandards White Paper, p. 30.
[11] “Neuroethics of Neuromarketing,” Emily R. Murphy, Judy Illes, and Peter B. Reiner, J Consumer Behav, July-October 2008, 293-302.