Hook-up, love affair or happy marriage?

Editor's note: Sean Green and Neil Holbert are with the University at Buffalo, the State University of New York. This article is adapted from their chapter “Neuroimaging and marketing research” in Leading Edge Marketing Research, edited by Robert J. Kaden, Gerald Linda and Melvin Prince. Reprinted by permission of SAGE Publications (www.sagepub.com). For a list of references and citations mentioned in the article, e-mail Joseph Rydholm at joe@quirks.com.

Neuroimaging represents a relatively recent new set of tools that promise to help marketing researchers solve one of the most intractable of problems: When respondents are asked questions, we cannot be sure the answers they provide are either true or accurate. Respondents misremember, they forget, they lie, they don’t even know about deep motivations for their behavior. All answers are intermediated by conscious and subconscious processes unknown and unknowable to researchers.

So for many years now, an ideal form of data has been sought, data that are unsullied by respondent-induced errors and biases. The mind and body of the respondent would produce this data automatically without being mediated by spoken or written answers. Earlier examples include pupilometrics, use of lie detectors, analysis of voice pitch and many others. Over time, these all foundered because, while responses to stimuli could readily be discerned, their consistent interpretation was missing, perhaps most particularly, their valence – whether something was liked or disliked.

Enter neuroimaging, borrowed from medical research, which not only produces automatic, unmediated responses but, according to its several practitioners, also allows clear interpretations of what the data mean.

But why do we need neuroimaging techniques at all? Simply put, in marketing research, we can get answers by asking questions (or, in some cases, by not asking questions). Yet what if the answers to those questions are wrong? What if respondents are bored by 25-minute surveys on the minutiae of their laundry habits and don’t really pay attention to the questions? What if people get in a yea-saying mode (and give positive answers because they figure that’s what is wanted); or in a I’m-not-a-dummy mode (and give answers, any answers, to questions they really don’t know anything about); or they just forget or misremember; or, toughest to handle of all, what if they get into a defensive-ego mode (where they decide that what they say about themselves is more important than anything they say about the matter being researched)?

All of this is possible in the realm of questioning and can lead us astray. Therefore, marketing researchers also use non-question-asking methods. Non-question-asking techniques can, broadly, include observation (seeing what people do, rather than what they say), including ideas from ethnography (what they do in a specific cultural context); diary research (don’t test people’s memories – ask them instead to record what they do as they go along); mystery shopping (our interviewers pretend to be shoppers); garbology (what do people throw out?); psychomechanical methods, where we use devices like eye-tracking machines (to see what parts of stimuli catch the respondent’s eye first, the trajectory that the eye follows and the time spent on each element of the scene); and tachistoscopes and computer displays that shed light on stimuli for various times to see how the participants react to brief glimpses of a display element.

While their responses may sometimes mislead or confuse us, in all of these techniques, respondents cannot really lie. That’s why we use them; it is good for us and also good for consumers, from whom we are getting information that can ultimately lead to products, services and issues that we hope will fill their needs.

See what happens to their brains

And now there’s neuroimaging, where we (literally or figuratively) attach wires to people’s heads and see what happens to their brains when they are exposed to stimuli or we put their heads in scanners to pick up the telltale radiation or fluctuations in magnetic force that accompany their thoughts, emotions and decisions.

In past decades, researchers have been able to use psychophysical and physiological methods to obtain general information about a consumer’s alertness (Krugman, 1971) emotional reaction to advertising or attention to particular elements of an ad. However, the ability to observe neural correlates of the decision process itself provides a new and conceptually different sort of information to marketing researchers. The possibility is that a researcher might be able to say to a respondent, “Your verbal responses say that you don’t like the product but your brain says that you do.”

Furthermore, neuroimaging might often provide insights not only into the consumer’s decision but also into how and why the consumer arrived at it. For example, neuroscientists can point to brain areas that are associated with reward, emotional influences on decision-making and evaluation of a product’s status or quality.

If marketers can follow the trail of neural breadcrumbs that leads from observing an ad, perceiving its contents, responding emotionally and cognitively and finally making the decision to buy, then this gives marketers the power to fine-tune their promotion so the consumer not only buys the product in the end but also buys it for an understood reason – as part of a long-term marketing strategy.

And this promise has never been offered before by a research tool.

'The map is not the territory'

But are neuroimaging techniques really windows into the mind? As Alfred Korzybski once said, “The map is not the territory” (Korzybski, 1948, p. xvii). Just as consumers cannot be reduced to the pie charts and bar graphs of a marketing researcher’s PowerPoint presentation, the entrancing images of an fMRI or PET scan are not snapshots of the mind itself. To put it another way, we may have taken the black box and painted it every color of the rainbow, but figuring out its secrets may be as challenging as ever.

Paradoxically, as marketing researchers are becoming increasingly aware of the potential of neuroimaging, a backlash – or at least a cautionary tug on the reins – has been building against the potential for erroneous or misleading interpretations of neuroimaging data. Some questions have risen in academia (Klein, 2010; Miller, 2008; Poldrack, 2006) and in the popular media (Horstman, 2010, pp. 80–83) and thus, neuromarketers should keep the following limitations in mind:

Brain scan images are not photographs. They are more like complex charts. The colors on a brain scan image do not arise directly from whatever signal (e.g., radio wave or gamma ray) the scanner is detecting. When neuroscientists analyze brain scan data, they carry out statistical tests to determine which color to paint each part of the image. Therefore, reading a brain scan requires as much care and skepticism as reading a bar graph of a survey result (Klein, 2010).

A brain scan image is, typically, created by comparing the experiment scan against a baseline or control. The brain is always doing more than one thing. If it weren’t, you’d stop breathing every time you got lost in thought. To account for this, techniques like fMRI, typically, require the researcher to subtract a baseline level of activity from the result. For example, a researcher might compare two conditions, one in which the participant looks at an advertisement and one in which the participant looks at a white screen. If the brain activity in these conditions is different, it does not necessarily mean that the advertisement content caused this difference. Perhaps the white screen was boring. Perhaps it hurt the participant’s eyes. The choice of a control or baseline condition is important to the meaning of the results.

Parts of the brain do more than one thing. We have already said that the brain can handle more than one task and that is true at the local level too. For instance, seeing the Coca-Cola logo before tasting the drink increases activity in the dorsolateral prefrontal cortex or dlPFC (McClure, Li, Tomlin, Cypert and Montague, 2004), linking the dlPFC to brand information. However, the dlPFC also responds in other contexts, such as word-recognition tasks (Fiebach, Ricker, Friederici and Jacobs, 2007). So what does it tell us when a company logo makes the dlPFC light up? It is important not to rely solely on labels like “reward center” or “judgment center” and instead to consider the range of possible reasons for a particular brain area’s response.

Neuroimaging studies are rarely done in a real-world setting. It would be fantastic if there were a lightweight, portable fMRI machine. Then we could just strap it to our participants’ heads and set them loose in a shopping mall. However, the current reality is that the kinds of techniques that let you pinpoint the location of brain activity (PET, fMRI and MEG) constrain the participants and, at the very least, provide a tangible reminder that they are in an experiment. At worst, they can deny a random sample when some participants don’t feel like sticking their heads in the scanner. That is not to say that experiments aren’t valuable but rather to suggest that experimental findings might be less applicable to situations where the consumer experience and the experimental participant experience are substantially different.

Most neuroimaging studies are conducted among small samples, usually in a single location. Forget about representative sampling among segments or geography or usage rates; forget about random sampling or statistically-reliable sampling (in the traditional sense). Typical studies employ a sample of just 20 respondents, which may be fine for medical research on brain function. It is less clear that these small samples will provide marketers what they seek and it is quite clear that they require marketers to take a leap of faith that the technology’s promise will overcome its current limitations in target audience representativeness.

Neuroimaging studies (so far) are very expensive and beyond the budgets of any but the largest of marketers.

So where does that leave neuromarketing? To say it succinctly, caveat emptor. Businesses aiming to use neuroscience to have an edge over their competition should be informed and discerning and know the right questions to ask when someone pitches them a marketing solution based on their forays into the black box. As Brainard (2008) notes, “There are many responses associated with each area of the brain, however, and as promising as imaging may be, it hasn’t come close to directly identifying the exact emotions in play” (para. 5).

In sum, there is the appeal of new technology, tempered by caution. As Northrop (24) so elegantly put it, “Men everywhere ... begin with the aesthetic emotional principle in the nature of things and come to the rational principle ... only later, if at all. One feels the beauty of the sunset before one learns of the internal constitution of the stars” (p. 62).

At a crossroads

Neuroimaging and the field of neuromarketing that has emerged from it are both in their infancy, yet already, they appear to be at a crossroads. If neuroimaging is seen as an invasive process that calls to mind a hospital, laboratory or interrogation room, it is likely to be viewed very warily and very much with a jaundiced eye. However, it has the potential to extend its influence well beyond science and theory and into practical day-to-day marketing.

So in all this hope of finding a Holy Grail into the consumer’s mind via neuroimaging, there are objections, limitations and caveats. Let’s recap them.

  1. At a very basic level, it can be argued that neuroimaging, such as psychoanalysis, is a medical matter, plain and simple, and results should be used only for medical purposes and not for selling things.
  2. Use of neuroimaging makes some ethically queasy. Although neuroimaging vendors claim to offer special insights into consumer behavior, do we really want to risk exploring the brains of those who are not necessarily that aware among adults – to say nothing of those with fewer defenses, such as children?
  3. There are many influencing factors on consumer behavior in general and on responses to research inquiries beyond what we as marketers proffer. These include influences from personality, family, friends, many groups, the culture, noncommercial messages of all sorts and just plain chance.
  4. There is limited validity to date showing that neuroimaging results relate to actual buying behavior in the marketplace. Studies involving wines, coffee and soft drinks are cited (Harris, 2008) but correlations and effects are limited.
  5. There is the potential for massive and dangerous oversimplification. As noted, the tools of neuroimaging are varied and the brain itself is frighteningly complex. To say it yet again, we are not dealing with bulbs on a Christmas tree that light up (or don’t) in any neat, clear or simple way.
  6. Will marketers (and the public) be able to distinguish real scientific breakthroughs from overblown claims? Will consultants employing exotic machines but very little real scientific knowledge try to entice businesses into spending real money on “junk” science? (After all, if X number of wires gives us useful information, why not 2X or 3X ... or nX? See Luck, 2005, pp. 122–124 for a technical discussion of using large numbers of electrodes in EEG recording.)
  7. Isn’t it possible that much of the appeal of neuroimaging as a contemporary hot marketing research tool comes from the fact that it is part of a techno-wave that has engulfed and enchanted a new generation of marketing researchers?

On the other hand, we may find that, over and above everything else, this dramatic new tool can bring forth fruits that may, by themselves, justify our attention to it. Thus, in methodology, “A key advantage of the use of imaging procedures in marketing is primarily the possibility of developing new theories about marketing-related behaviors through the use of the inductive method” (Kenning, Plassmann and Ahlert, 2007, p. 66; translated by the authors with the aid of Google Translate). And in the realm of business itself: “[Marketing] researchers have an unparalleled opportunity to adopt cognitive neuroscientific techniques [to] ... significantly redefine the field and possibly even cause substantial dislocations in business models” (Lee, Senior, Butler and Fuchs, 2009).

Love affair or lasting marriage?

So are neuroimaging and marketing research a hook-up, a passing love affair or a lasting marriage? We like to quote Dan Ariely, Duke University, and Gregory S. Berns, Emory University:

“It is too early to tell, but, optimists as we are, we think that there is much that neuromarketing can contribute to the interface between people and businesses, and – in doing so – foster a more human-compatible design of the products around us. At the same time, neuromarketing as an enterprise runs the risk of quickly becoming yesterday’s fad. ... If we take neuromarketing as the examination of the neural activities that underlie the daily activities related to people, products and marketing, this could become a useful and interesting path for academic research, and, at the same time, provide useful inputs to marketers.” (Ariely and Berns, as quoted in Hernandez, 2010, p. 8)

Alternatively, in the also-developing field of neuroethics, researchers have already raised concerns about the use of neuroimaging technology to exploit vulnerable populations and undermine free will (Farah, 2005).

And tensions have risen between entrepreneurs using neuroscience techniques commercially and researchers who question the foundation of their work. For example, the “Farwell Brain Fingerprinting” technique for applying neuromeasurement to criminal investigation has faced scrutiny (Rosenfeld, 2005) and a neurologist has questioned Dr. Daniel Amen’s use of brain scans in the diagnosis of psychological disorders (Burton, 2008).

The perception of neuromarketing as trustworthy and viable rests not on any particular article or controversy or marketing initiative but on how it comes to be viewed by the three spheres: marketing, consumer and scientific. If scientists reject neuromarketing as pseudoscience; if marketers come to view it as unproven, gimmicky, or too complicated or expensive to be useful; or if the public grows defensive or suspicious of neuromarketing research, it may not take root, irrespective of its scientific merit, its utility for businesses or its potential to benefit consumers.

It doesn’t have to be that way. In fact, it could add richness and promote ethical choices rather than discourage them. It is possible that neuroimaging (and neuromarketing) could, according to Lee, Broderick, and Chamberlain (2007), “contribute to marketing ethics in many ways ... (and) research into advertising effectiveness ... can contribute more than just finding ... (that) ‘buy button’ in the brain. In fact, exploring exactly what elements of an advertisement are critical to awareness (of), attitudes (towards) and evaluations of products ... should reduce firms’ reliance on the ‘blunt instruments’ of blanket coverage, shock tactics, or sexual imagery.” (p. 203)

No doubt the excitement and controversy surrounding neuroimaging and neuromarketing – reading my brain and then leading me into buying temptation – will continue. As Farah (2005) puts it, “The question is, therefore, not whether, but rather when and how, neuroscience will shape our future” (p. 39).

Do we, finally, hear a call for business pragmatism as we look into the matter? Here’s one from de Abaitua (2009):

“While some are attracted to neuromarketing’s promise of monitoring the emotional responses of an individual’s brain, other experts feel that the answer to emotional engagement lies ... in the domain of signs and symbols, known as semiotic analysis. Many brand managers will engage a grab bag of these feuding experts and pick and mix from their advice. The various sects and dogmas of marketing are treated expediently by their clients ... they just want an approach that works.” (para. 3)

Or, to paraphrase the Scarecrow in The Wizard of Oz, we could really move ahead in our science and in our marketing “if we only had the brain.”

Maybe we do.