Neuroscience-Inspired Design: From Academic Neuromarketing to Commercially Relevant Research

On the Power of Persuasive Brain Images and Neuroscience Explanations

While it is often suggested that neuroimaging techniques such as fMRI help provide answers to specific business questions, colorful and scientific-looking brain images also serve a purpose in their own right within organizations as a powerful tool of persuasion. This is true regardless of the scientific value, or meaning, that such images may have. In fact, one often sees competing factions within an organization battling it out in terms of the most appropriate direction to take (for a new product launch, say). Those wielding the brain scans are often at something of an advantage relative to other groups brandishing more traditional forms of research data (think questionnaire, or focus group research; Catterall & Maclaran, 2006; Lunt, 1981; cf. O’Connor, Rees, & Joffe, 2012).

An oft-cited study by McCabe and Castel (2008) hints at just how persuasive brain images can be. These researchers demonstrated that even scientifically literate audiences (who should, presumably, know better) are more likely to believe an argument presented in a news story if it happens to be accompanied by a colorful brain-scan. Worryingly, this was true even if the brain image had nothing to do with the argument that was being made! That said, more recent research shows that claims regarding the seductive appeal of the colorful brain-scan may not be quite as strong as McCabe and Castel’s early findings led many to believe (see Farah & Hook, 2013; Michael, Newman, Vuorre, Cumming, & Garry, 2013).

Elsewhere though, Weisberg, Keil, Goodstein, Rawson, and Gray (2008) have demonstrated that arguments that have been given a neuroscience explanation (or spin) are judged more plausible than those that are not by novices and students alike. As O’Connor et al. (2012, p. 220) note, “Logically irrelevant neuroscience information imbues an argument with authoritative, scientific credibility.” No wonder, then, the growing presence of neuroscience in the public domain. Worryingly, according to an analysis of almost 3,000 neuroscience articles that appeared in the British press in the opening decade of the 21st century, media reports often distort and embellish the findings described in the underlying scientific research (O’Connor et al., 2012).

Given such findings, it should come as little surprise to find that critical appraisals of what neuroscience really tells us have also become more frequent. Some of the overstated claims of those who use neuroscience, and neuroscience-based explanations in the public domain, to theorize well beyond their own areas of expertise has recently come in for some vigorous, and well-deserved, scrutiny/attack (see, e.g., Editorial, 2004a; Marcus, 2012; Quart, 2012; though see also Brammer, 2004). This backlash against “brain porn,” defined by Alissa Quart (2012) as the “willingness to accept seemingly neuroscientific explanations for, well, nearly everything,” has taken place across a range of disciplines from politics to economics, and from literary theory through to neuroaesthetics. One particularly bruising public spat occurred when a group of neuroscientists and neuromarketers published the results of an fMRI study of a small number of “swing voters” in the 2008 U.S. elections as they responded to the leading presidential candidates (see Iacaboni et al., 2007). Scathing critiques of this research, which appeared in an op-ed (invited opinion) piece in the New York Times, were not long in coming. Some went so far as to compare it to modern-day phrenology (see Aron et al., 2007; Editorial, 2007).

On the Problems of Commercial Neuromarketing

It is important to note that the new techniques of analyzing the mind of the customer, consumer, or whoever else the target audience of a company or organization might happen to be have their own constraints and limitations (see, e.g., Javor, Koller, Lee, Chamberlain, & Ransmayr, 2013; Lee, Senior, & Butler, 2012a, 2012b; Senior & Lee, 2008, 2013; Senior, Lee, & Butler, 2011, for detailed discussion of a number of these points). Some of the most salient of which include the not-inconsiderable cost (Ariely & Berns, 2010), the limited availability of scanners in many parts of the world (especially in the developing world), and the banning of commercial neuromarketing in certain countries (see Oullier, 2012; see also Eaton & Illes, 2007; Editorial, 2004b; Murphy, Illes, & Reiner, 2008). There is also a question-mark when it comes to the ecological validity of isolated participants lying flat on their back, often with their head restrained, while subjected to loud background noise, inserted halfway down a narrow tube (e.g., Spence & Piqueras-Fiszman, 2014; Wall, 2013).

Another potential concern with many of the commercial neuromarketing studies that have been conducted to date concerns just how representative of society at large the small number of participants who typically take part in neuroscience research really are (though, of course, a similar criticism can be leveled at much of the focus group research too). A few years ago, for instance, Henrich, Heine, and Norenzayan (2010) highlighted how the majority of participants in psychological research are from societies that are WEIRD—that is, Western, educated, industrialized, rich, and democratic. Worse still, it has been estimated that somewhere in the region of 95% of all psychology research is conducted on North American psychology undergraduates (see Jones, 2010)—hardly the most representative sample, then, when it comes to marketing research (or any other kind of research for that matter). There would seem to be little reason to believe that the makeup of those taking part in neuroimaging experiments should necessarily be any different. Finally, there is also the potential public/media backlash to consider. Scare stories concerning big corporations prying into the minds of their consumers are never too far from the media’s gaze. While such a putatively negative response is by no means restricted to the field of commercial neuromarketing (see, e.g., Packard, 1957, for an early scare along these lines), the headlines suggesting that companies are peering inside their consumers’ brain to try and find the “buy button” typically don’t go down all that well in the boardrooms of the companies that have been ‘fingered’. That said, corporate concerns regarding the reputational risk associated with this kind of criticism would seem to have declined somewhat in recent years.

More problematic than any of the above concerns, though, when it comes to commercial neuromarketing is the fact that, practically-speaking, the time-frame for conducting neuroimaging research often falls well outside of the tight deadlines that those working in the fields of marketing and product development typically have to operate within. Furthermore, while neuroimaging can provide a relatively clear answer to black-and-white questions, it is rather less well suited to discriminating between shades of gray questions. And, very often, it is just those shades of gray questions that the marketing manager or product developer desperately wants an answer to. Given such problems, business leaders might well be forgiven for wondering whether the neuroimaging approach is really worth the effort (not to mention time and cost). One other worry here is the exceedingly small sample sizes that are sometimes used in published neuromarketing research (e.g., 3 participants in Ambler, Ioannides, & Rose, 2000; 11 in Stoll, Baecke, & Kenning, 2008; and just 14 in Basso et al., 2014). This problem obviously being exacerbated by the fact that those participants were, anyway, most likely WEIRDos (Henrich et al., 2010; Jones, 2010).

Over the years, one of the problems with many neuromarketing companies is that they have tended to make claims that are hard to evaluate independently. Too often, they offer their clients “black box” solutions. “Go with solution B, not A or C,” they say. When clients ask why, they will often be told that that is confidential information and hence that the methods used cannot be divulged. As Wall (2013) notes, “The tendency of these companies to keep their methods secret also hampers serious evaluation.” This would not necessarily be such a problem if it weren’t for the fact that so many spurious claims are made on the basis of commercial neuromarketing techniques/approaches that simply could not give a reliable answer to the question posed (see Wall, 2013). As Oullier (2012) notes, “That a technique is still in its infancy doesn’t stop people from trying to sell it or from buying it. And damage can be done even if the victims of neuromarketing hype are not the general public but the gullible heads of companies who are being overcharged.”

There have also been a number of well-documented cases where the commercial neuromarketers have gone beyond the conclusions that can legitimately be drawn from the data. As Eric du Plessis (2011, p. 151) said of the sensationalist tone used to describe the findings of a $7 million study reported by Martin Lindstrom in his 2008 book Buy-ology: How Everything We Believe About Why We Buy Is Wrong, in which more than 2,000 people had their brains scanned: “I have my doubts when any study concludes that neuroscience has now proved that everything we knew about behaviour is wrong.” (Lindstrom has also come in for some robust criticism from other quarters too; see, e.g., http://www.russpoldrack.org/2011/10/nyt-editorial-fmri-complete-crap.html.) Sometimes there is also a question-mark as to just how similar the situations and decisions that are modeled in the scanner environment are to those of real life (Anonymous, 2005). For instance, having the participants in one’s neuroimaging experiment make a judgment about which of a range of cars they see in pictures on a computer screen “that they would buy” (Tusche, Bode, & Haynes, 2010) is obviously very far removed from someone actually going out to spend his or her hard-earned cash on a new vehicle (cf. Ariely & Berns, 2010).

On the Need for New Solutions

Given these problems with commercial neuromarketing, one might well start to wonder why the neuroscience approach to the study of consumer behavior has become so popular in recent years. There are several answers here. On one hand, across a range of sectors, the majority of the estimated 30,000 new consumer product launches each year fail. The actual figure is variously estimated to be around 70% to 80% (from market research performed in 1997 by Linton, Matysiak, & Wilkes Inc.; Crawford, 1977; Underhill, 1999, p. 163) all the way up to 95% (see, e.g., Nobel, 2011; Schneider & Hall, 2011; see also Robinson, 1998, p. 134; Zaltman, 2003, p. 3). As such, current/traditional methods clearly are not working. This is all the more worrying given the amount of money that is being spent. As Mast and Zaltman (2005, p. 423) note,

In 2002, an estimated US$ 6.8 billion was spent on conventional marketing research tools in the United States alone. This is an enormous amount of money considering that there is little scientific evidence to support the widespread use of focus groups and growing concerns about the accuracy of other verbal report techniques.

Of course, every new marketing technique comes with the promise that it will help bring down the number of new product failures. However, while it is possible to find in the literature post hoc accounts of how commercial neuromarketing can help explain what has already happened, it is much harder to find examples of future successes being predicted in advance using such techniques (though see Jarrett, 2015). One explanation for this might be because such examples are kept under lock-and-key by the companies who funded the research in the first place. Alternatively, the scientific research journals simply might not find there to be enough “theoretical meat” on those articles describing research that is too close to the real concerns of the marketplace. Despite all of the hype, there really haven’t been all that many examples where the commercial neuromarketing (or consumer neuroscience) approach has led to long-term market success—success that could not have been predicted by behavioral techniques. One of the few examples that gets closest to what we are after here, by Gregory Berns and his colleagues will be described later (Berns, Capra, Moore, & Noussair, 2010; Berns & Moore, 2012).

Please do not get me wrong. I am not an evangelistic neuroskeptic (see, e.g., Power, 2008; Tallis, 2008a, 2008b), although I am sometimes wheeled out to play that role (Spence, 2015). Rather, I am a strong proponent of the “neuroscience-inspired approach to multisensory design” (Spence, 2012d; Wethey, 2013). It is clear to most people, I think, that neuroscience has made phenomenal progress in furthering our understanding the fundamental workings of the human mind in recent decades. What is more, neuroscience (here I am thinking specifically of academic neuromarketing research) ¹ has delivered, and continues to deliver, worthwhile understanding of the basic mechanisms underlying many of our real-world behaviors (see Ariely & Berns, 2010; Yoon et al., 2012, for reviews). However, from a practical point of view, one should always ask whether neuroimaging solutions per se are strictly necessary, appropriate, or even practical. In the majority of commercial settings, the answer may well be no (see also Page & Raymond, 2007). Some combination of the various limitations highlighted above often mean that neuroimaging is not the key technique in problem solving or new product development, at least not in the commercial setting.

Neuroscience-Inspired Design

At one level, neuroscience research (what I will call academic neuromarketing) undoubtedly plays an important role in theory development in the area of consumer decision making (e.g., Ariely & Berns, 2010; Yoon et al., 2012). Ultimately, though, every company or organization is really interested in monitoring, predicting, and/or changing the behavior of their target audience, be they customers, consumers, the general public, or even other companies when it comes to business-to-business interactions. Therefore the question that anyone who is toying with the idea of using neuroscience to try and understand the behavior of their target audience should be asking themselves is whether that behavior change can’t simply be measured more effectively, and possibly more cheaply, by actually observing some aspect of people’s behavior or assessing their opinions. Sure, people have problems verbalizing what they think, but a range of implicit behavioral tests are now available (Mast & Zaltman, 2005; Parise & Spence, 2012). More often than many companies realize, the most efficient solution may result from neuroscience-inspired behavioral testing rather than commercial neuromarketing. If what you are really interested in is people’s behavior, and how to navigate behavior change more effectively, then surely you ought to be able to achieve that by studying behavior in the first place (cf. Page & Raymond, 2007). What relevant additional information exactly could knowing what is going on in the brain provide? Where neuroscience has provided useful insights it has normally come from relying on the research findings that are already out there in the public domain (in some cases, from published academic neuromarketing research), rather than on a particular company commissioning their own research from scratch. Note that the distinction (and gap) between academic and applied practice in the neurosciences has also been highlighted by Senior and Lee (2013).

Mast and Zaltman (2005, p. 426) are worth quoting at length here:

The use of neuroimaging techniques in marketing contexts is very much in its infancy. Although very promising, a great deal of experimental work is required to determine what set of problems are best addressed using this approach and how neuroimaging techniques might best supplement existing methods. Neuroimaging does not generate hypotheses and to formulate interesting hypotheses, it is necessary to have a deep understanding of the issues involved in the research context. For this reason, it is likely that neuroimaging methods might best be used in areas where there is substantial knowledge already. ²

Ultimately, for anyone interested in understanding, predicting, and changing behavior, what is of interest must ultimately be observable in terms of people’s behavior (Breiter et al., 2015; Falk et al., 2016). The majority of our work is inspired by the latest findings from cognitive neuroscience, although the research that we do is mostly behavioral in nature, often utilizing some form of implicit behavioral testing. The approach is perhaps best characterized as neuroscience-inspired design. In other words, the research from the cognitive neurosciences, broadly defined, undoubtedly provides a host of useful ideas, insights, suggestions, theories, and methods that can often be fruitfully incorporated into the design of the study when one is trying to address a variety of real-world problems. What the cognitive neurosciences offer is really a range of cutting-edge techniques that have been finely honed to detect often-small behavioral effects (or differences) as robustly and efficiently as possible. As such, these techniques can play an important role in claims support. (Though note that such research rarely makes it into the journals.) The neuroscience not only provides techniques but can also provide relevant fundamental insights too (particularly in the area of academic neuromarketing).

Interim Summary

To summarize, the solution to business problems very often involves seeking inspiration and insight from neuroscience research that has typically been conducted by others (often, but not always, this may be best characterized as academic neuromarketing). The findings of such research is then used to feed into the design of (neuroscience-inspired) behavioral tests that can be conducted within the relevant timeframe and budgetary constraints. This kind of solution would appear more valuable than that offered by the majority of commercial neuromarketing practitioners.

In the remainder of this article, we take a critical look at the growing interest in, and usefulness of, research on the neuroscience of behavior (i.e., understanding how the brain does it) from a practical commercial point of view. Three real-world areas from the neuroscience of behavior/perception are highlighted: neuroergonomics (Fafrowicz & Marek, 2007; Parasuraman, 2003; Spence, 2012b), neuromarketing (see Breiter et al., 2015; Legrenzi & Umiltà, 2011; Spence & Gallace, 2011; Spence, Puccinelli, Grewal, & Roggeveen, 2014), and neurogastronomy (Shepherd, 2006, 2012; Spence, 2012a, 2015). By the end, the benefits of the neuroscience-inspired approach to design and, by extension, to the modification of people’s behavior should hopefully be clear. Ultimately, knowing about the mind of one’s customer or client has got to provide a competitive advantage in the marketplace. While the nature of that advantage is currently easier to see in certain sectors than in others, eventually, it will be true across the board. ³

Neuroergonomics: On the Neuroscience-Inspired Approach to Interface Design

The term neuroergonomics refers to the study of the neural circuits involved in people’s interactions with interfaces (Fafrowicz & Marek, 2007; Parasuraman, 2003). The majority of the research in this area has been in the domain of driving (see Spence, 2012b, for a review). Indeed, in recent decades, researchers have started to investigate the network of brain areas that “light-up” while people are driving or, more usually, while they are engaged in some form of simulated driving task (i.e., when driving in a video game/simulation; see Calhoun et al., 2002; Spiers & Maguire, 2007). One of the hopes of those working in this area has been that by gaining a better understanding of the mental state of the driver, it might be possible to modulate the delivery of warning signals, and other in-car alerts/information. So, for example, a drowsy older driver may well need a greater lead time than an attentive younger driver when responding to a warning signal. One of the key challenges here, at least while people are still driving cars (Markoff, 2011), is to make sure that any warning signals are presented with enough of a lead in order to allow the driver to update their behavior, but not so far ahead of a potential danger that the warning signals give rise to “cry wolf” phenomenon (see Ho & Spence, 2008). There is a constant arms race between the delivery of ever more technology that may potentially distract the driver, and the growing use of new technologies to warn of the danger of impending danger on the road ahead (e.g., Ashley, 2001; Hudson, 2013; Vlasic, 2008).

Many of the car companies have been interested in analyzing the neural signatures associated with specific driving behaviors. Some, for instance, have investigated how much useful information can be derived from drivers wearing a cap to measure their electrophysiologcal responses while on the road (e.g., Anonymous, 2005; Harlow, 2011). Meanwhile, a helmet developed by Honda a few years ago allowed drivers to execute one of four commands using direct neural control (Harlow, 2011). The difference here from commercial neuromarketing, say, is that the large sums of money involved, and the relatively slow rate of progress, are not necessarily such limiting factors when it comes to research in the field of neuroergonomics. That is, the car companies are used to spending huge sums on research and the timelines tend not to be as punishing as in the world of marketing.

A few years ago, for instance, Toyota wanted to know whether it was possible to “demonstrate that cognitive neuroscience can help to design multisensory warning signals for drivers that are significantly better than a smart (i.e., intuitive) engineer can come up with.” The inspiration for the research addressing this question came from neurophysiological findings reported at the single-cell level (Stein & Meredith, 1993). In particular, key neuroscience rules underpinning multisensory integration that have been incorporated into recommendations for the design of more effective warning signals include everything from the beneficial effects of spatial colocation (e.g., Ho & Spence, 2008; Spence & Driver, 2004; see also Liu & Jhuang, 2012), temporal synchrony, or perhaps slight asynchrony (e.g., Chan & Chan, 2006; Spence & Driver, 1999). There has also been interest in the breakthrough of multisensory (as compared to unisensory) warning signals, at least when the component unisensory signals happen to be presented from the same spatial location, or direction, at more or less the same time (see Spence, 2010a, for a review). Importantly, none of these principles/solutions would appear to have been picked-up intuitively by the engineers working on the design of warning signals (Spence & Ho, 2008; Spence & Read, 2003).

Another relevant insight emerging from the basic neuroscience research conducted over the last decade or so relates to the discovery of specialized brain circuits that monitor the space just behind the head (see Occelli, Spence, & Zampini, 2011, for a review): This is the space that we do not see, at least not directly (and that we rarely think about). Interestingly, researchers have demonstrated that placing an auditory warning signal in this region of near-rear peripersonal space results in faster head-turning responses by drivers in the laboratory and also, more importantly, in the driving simulator than when the same or other modality warning signal is presented in frontal space (Ho & Spence, 2009; Spence, 2012b). Note that this counts as an example of neuroscience-inspired design: That is, the discovery by neuroscientists of these specialized brain circuits preceded, and subsequently prompted, the applied behavioral research. It would not have been practical, or, for that matter, financially viable, for the car company concerned to pay for the underlying brain research in the search for such neural circuits that could one day feed into the design of enhanced warning signals for car drivers. ⁴ However, it was eminently sensible for the applied researchers to piggyback on the emerging neuroscience insights that were appearing in the scientific journals. The practical challenge (this time for the engineer) was to try and figure out how those findings could be used to aid in the development of more efficient warning signals. In this case, Toyota were unable to capitalize on the insights by developing a new piece of headrest-mounted technology that they could protect, and by so doing, protect the bit of space behind the driver’s head that our research had highlighted to be so interesting. Nevertheless, other vehicle manufacturers (e.g., Volvo) have now started to present auditory information via the headrest (see http://apps.volvotrucks.com/launch/fhfm/en-gb/downloads/FH_image.pdf, p. 27).

Over the past decade or so, the neuroscience-inspired approach has also been applied to the design of a variety of other user interfaces (Baldwin et al., 2012), including those used by pilots (Previc, 2000) and those used in air-traffic control (Ngo, Pierce, & Spence, 2012). One day the same neuroscience-inspired insights might well be applied to enable traders to respond more rapidly on the trading floor too. What all of these various situations have in common is that every millisecond counts: Just take the following quote from an article that appeared in the magazine Information Week: “A 1-millisecond advantage in trading applications can be worth $100 million a year to a major brokerage firm” (Martin, 2007). Hence, in conclusion, it can be argued that knowing about the mind of the interface operator, no matter whether the operator be on the road, in the air, or on the trading floor, can potentially lead to enhanced neuroscience-inspired design solutions (see Spence, 2012b). ⁵

Neuromarketing: Sensory Marketing and Multisensory Product Design

The number of commercial neuromarketing volumes out there on the marketing shelves has grown rapidly in recent years (see, e.g., Lindstrom, 2008; Pradeep, 2010; Renvoisé & Morin, 2007). A quick look at the back-covers of these books reveals that the authors promise revolutionary new techniques that “speak to the old brain” (Renvoisé & Morin, 2007), show the reader how to press the “buy button” in the customer’s brain (Lindstrom, 2008; Renvoisé & Morin, 2007; though see also Blakeslee, 2004; Hubert & Kenning, 2008; Lee, Broderick, & Chamberlain, 2007), or else reveal the secrets for selling to the subconscious mind (Pradeep, 2010). Or take the following from a Millward Brown study conducted on behalf of the Royal Mail that apparently showed that physical media (i.e., as opposed to digital) generated “deeper” brain activity (if only one knew what that means, or whether or not it is a good thing!; Millward Brown, 2009). There can be no doubt though that consumer neuroscience is big business (Hubert & Kenning, 2008). Nowadays, the term neuromarketing itself covers a number of distinct areas of research, including the study of advertising effectiveness, product design (what some call neurodesign; Legrenzi & Umiltà, 2011), neuromerchandizing (Roper, 2014), and so on (see Breiter et al., 2015). In the following sections, we will briefly take a look at a few of these areas.

The Era of Predictive Neuromarketing

A couple of recent studies have demonstrated that brain scans can be used to help predict future consumer behavior. In one, Falk et al. (2016) demonstrated that neuroimaging data could be used to help predict whether those reading an antismoking email would end up clicking on a link that took them to a website offering advice on quitting or not. That is, neural evidence obtained in a small group of participants (N = 50), specifically activation in a “self-localizer” defined region within mPFC was capable of predicting the population response (N = 400,000 individuals who received an email). In the other study, neuroimaging helped predict the box office success of movies (Boksem & Smidts, 2015). Although both studies were published retrospectively (i.e., once the outcome was already known), taken together they do start to hint at how neuroimaging evidence might one day be used in a genuinely predictive manner when trying to assess the likely success or failure of a planned intervention in the marketplace.

Crucially, Falk et al. (2016) demonstrated that by adding in the neuroimaging data they were able to explain 65% of the variability in the results as compared to just 38% of variability in terms of people’s (N = 63) click-through rates on the emails being explained by the results of an MTurk online survey. Assuming that the right questions were asked in the MTurk survey then such neuroimaging results are potentially exciting. Note, though, how it is the converging methodologies approach that proved most useful here. That is, no one is suggesting, at least not yet, that neuroimaging will completely replace other behavioral techniques. Rather the suggestion is that under the appropriate conditions it can provide a useful source of additional data.

Meanwhile, in the other study, Boksem and Smidts (2015) had 29 participants view a number of movie trailers for films that they hadn’t yet seen. They rated how much they thought that they would like the films and how much they would have been willing to pay (WTP) for the DVD. EEG was also recorded. Analysis of the WTP data helped to explain only 1% of difference between actual box office receipts of the best- and worse-performing films. However, adding data from the modulation of gamma brain waves (that seem to correlate with enjoyment) tripled how much of the real-world data could be predicted. Now, of course, here one can always wonder as to whether WTP is really the most appropriate behavioral measure, given that good/successful movies don’t cost any more at box office. The authors argued that while predicting only 3% of the variance might not sound like a lot, given the vast sums involved, every little helps. One final point to note about both of these studies is the modest sample sizes used (N = 50 in Falk et al., 2016; N = 29 in Boksem & Smidts, 2015) as compared to what one normally sees in audience/consumer research.

Neurogastronomy

Neurogastronomy is the term used to refer to the brain on food (e.g., Shepherd, 2006, 2012; Spence, 2012a; Toepel et al., 2015). As yet, the majority of the published research in this area has tended to be at the more academic rather than commercial end of the neuromarketing spectrum—what Javor et al. (2013) call consumer neuroscience rather than neuromarketing. There are undoubtedly a number of important questions in this field: everything from assessing the impact of branding, pricing, organic labels, and descriptive labels on people’s response to foods and beverages through to a consideration of what, if anything is different about the brains of obese individuals (or others suffering from some form of eating disorder). It is certainly clear that we need new ideas when it comes to trying to understand why society is currently having so little success in fighting the global obesity crisis. Indeed, as Shepherd (2012, p. 233) puts it, “A possible advantage of a new term like neurogastronomy is that it can help focus public policy more effectively in applying advances in brain science to issues related to food and flavour.”

It has been known for years that branding, pricing, and other forms of product-extrinsic information changes what people say about what they taste (see Davis, 1987; Martin, 1990; Spence, 2010b). What has never been clear, though, at least not until recently, is whether such information really changed the taste of the food, or merely what people chose to say about the taste. The classic experiment on the neural substrates of branding was conducted by McClure et al. (2004). The participants in this oft-cited academic neuromarketing study had their brains scanned while a cola drink (either Coke or Pepsi) was periodically squirted in their mouth while different visual information concerning the brand of cola that they were apparently tasting was projected onto a screen. Under blind tasting conditions, ventromedial prefrontal cortex activity was found to correlate with the participants’ behavioral preference. Qualitatively different patterns of brain activation were, however, observed depending on which brand of cola the participants thought that they were tasting. On being led to believe that they were tasting Coke, increased activation was observed in the hippocampus, dorsolateral prefrontal cortex, and midbrain.

In a more recent follow-up, Kühn and Gallinat (2013) observed greater activation in the left ventral striatum when their participants believed that the cola that they were tasting was a strong rather than a weak brand (e.g., Coke or Pepsi versus a national or fictitious brand). Intriguingly, this effect was stronger in those participants who drank cola infrequently, possibly pointing to a greater reliance on brand cues in less experienced consumers (see Plassmann et al., 2012, for a review of the consumer neuroscience literature on branding). Meanwhile, Linder et al. (2010) have demonstrated that labeling a food as organic led to increased activity in the ventral striatum, a part of the brain that is involved in controlling the motivation to acquire and eat food.

Neuroscientists working over in California investigated what would happen in the brain of social wine drinkers (students) when they were given different, and sometimes misleading, information about the price of red wine (Plassmann et al., 2008). A $5 bottle of wine was either correctly described in terms of its price, or else was mislabeled as costing $45 a bottle. Another bottle of wine actually cost $90 and was either described as a $10 wine or as a $90 wine. The third wine was correctly referred to as a $35 wine. The price was displayed on a computer monitor whenever a small amount of wine was squirted into the participant’s mouth. On some trials, the participants had to rate the intensity of the wine’s taste on a 6-point scale, whilst on other trials they rated its pleasantness instead. Sometimes, no behavioral response was required. While flavor intensity was not influenced by the price information, participants reported liking the wine more when labeled as expensive than when labeled as a cheaper wine (see also Plassmann & Weber, 2015).

Crucially, analysis of participants’ brain scans revealed increased blood flow in the mOFC in those trials in which a higher wine price was shown. While such results are clearly of theoretical import when it comes to understanding how product-extrinsic cues change the consumer’s response to food and beverages, it is perhaps less clear as to whether any company would really be willing to pay for the results of a study such as that conducted by McClure et al. (2004). What these and many other neuroimaging studies clearly do show, though, is that branding, labeling, and pricing information can give rise to changes in neural activation in some of the earliest brain areas coding the sensory-discriminative and hedonic aspects of gustatory, olfactory, and flavor information (see Spence, 2015, for a review). Undoubtedly, such results are of interest to the research scientist; what is less clear, is whether such results would satisfy (i.e., count as value for money) the food and beverage companies concerned. (Once again, we are back to the distinction between academic and commercial neuromarketing.)

I would like to end this section by looking at the results of a few of the studies that have addressed the question of how the brain of those with some form of eating disorder (such as, for example obesity) may differ when presented with (normally visual food cues) as compared to the brain of a normal weight individual (see Petit, Basso, Huguet, Plassmann, & Oullier, 2011; Spence, Okajima, Cheok, Petit, & Michel, 2015). For example, those individuals who suffer from binge-eating disorder and bulimia exhibit greater reward sensitivity, brain activation, and arousal, in response to viewing images of pleasant foods (Schienle, Schäfer, Hermann, & Vaitl, 2009). By contrast, obese individuals exhibit significantly less activation of the reward-related brain areas in response to food consumption than do normal weight individuals. Such a pattern of results suggests that those who are obese may anticipate more reward from food intake while, at the same time, experiencing less sensory pleasure from eating (Stice, Spoor, Bohon, Veldhuizen, & Small, 2008). In other research, obese individuals exhibited increased visual attention to food images as compared to normal weight individuals (e.g., Castellanos et al., 2009; see also Spence et al., 2015, for a review; see also Petit et al., 2016). Obese individuals also appear to be more sensitive to the expected pleasure of consumption than are nonobese individuals (Berridge, Ho, Richard, & DiFeliceantonio, 2010; Petit et al., 2011; Pursey et al., 2014). ⁸ In conclusion, academic neurogastronomy research has already provided a number of intriguing insights that will, no doubt, soon be picked up by commercial neuromarketing, if they haven’t been picked up already!

Conclusions

In conclusion, it is time to return to the question of whether the neuroscience approach to understanding the mind and behavior of the consumer really lives up to the hype (see Legrenzi & Umiltà, 2011). Under a small subset of conditions, it does. That said, it is fair to say that the excitement over the neuroimaging approach (specifically commercial neuromarketing) seen a decade ago has waned somewhat in recent years. There are clearly certain questions that behavioral techniques, even neuro-inspired behavioral techniques, simply cannot answer. The fundamental problems limiting the more widespread uptake of neuroimaging techniques by business (what has been labeled here as commercial neuromarketing) includes the time required to reach an answer, the expense involved, and the sheer number of participants that may be required to answer the shades of gray questions. There are also more mundane, but no less limiting day-to-day issues concerning the legality and ethics of commercial neuromarketing (see Oullier, 2012; Stanton, Sinnott-Armstrong, & Huettel, 2016). Furthermore, the lack of widespread access to the latest neuroimaging techniques in many emerging economies is also currently a major constraint for companies around the world. Finally, there is also very real potential for consumer/press backlash (see, e.g., http://www.dailymotion.com/video/xr6cos_neuromarketing-votre-cerveau-les-interesse-1-2_news for one media sting that caught up a number of those working in the field a few years ago). Given the potential negative press coverage, it should not come as any surprise that many of the companies who utilize commercial neuromarketing have been understandably reticent to have that information out there in the public domain.

Looking forward, excitement surrounds the possible use of neuroimaging to segment groups of consumers based on their differing patterns of brain activation (see Plassmann & Weber, 2015; Yoon et al., 2012). The hope is that one day, neuroimaging techniques could be used to account for some of the heterogeneity of consumer behavior that is so often seen in the marketplace. In terms of the implications for marketing practice, it is worth noting that commercial neuroimaging is getting cheaper, neuroimaging facilities are slowly becoming more widely available, statistical analysis techniques are continually improving, and the body of background knowledge (largely from published academic neuromarketing, or consumer neuroscience research) ever larger and more sophisticated. All of this ought to suggest a growing utility of commercial neuromarketing in the future (e.g., Smidts et al., 2014).

While neuroscience has undoubtedly made great steps forward in recent years (Ariely & Berns, 2010), the interface with business, and the concerns of big business continues to be challenging. The neuroscience-inspired (multisensory design) approach constitutes one of the more fruitful approaches in this area. By adapting the tools, techniques, methods, and understandings that have emerged, and continue to emerge, from the field of academic cognitive neuroscience research, broadly defined, it is possible to help address real-world business problems concerning people’s behavior, in a realistic timeframe, at an affordable cost, and ideally in a manner that is scalable. As Ariely and Berns (2010) note, “If neuromarketing is to compete with conventional marketing approaches on the basis of efficiency,” in the years ahead “then the costs of labour and overheads will have to be reduced.” Elsewhere, they state that, “It is not yet clear whether neuroimaging provides better data [than] other marketing methods” (p. 287). The same basic approach can be used no matter whether we are dealing with the neuroscience-inspired design of multisensory warning signals for car drivers (Ho & Spence, 2009; Spence, 2012b), trying to advise companies on the future design of their product packaging (Juravle et al., 2015; Parise & Spence, 2012; Spence & Zampini, 2007), or considering the design of more effective/healthy foods (Spence et al., 2015).

Ultimately, perhaps the most parsimonious line here is not to frame consumer neuroscience as a fundamental challenge to more traditional behavioral research methods, but rather as a complementary approach (Hubert & Kenning, 2008). Navigating the converging methodologies approach to applied research (be it in the area of organizational research, or elsewhere; see Becker & Cropanzano, 2010; Halvorson & Rock, 2015; Lee & Chamberlain, 2007; Powell & Puccinelli, 2012; Rock & Schwartz, 2007), and bringing traditional approaches to analyzing consumer behavior into the 21st century is a challenge that will likely keep many of those working in the field busy in the years to come. The argument put forward here is that we will likely make most robust progress if we keep the distinction between academic and commercial neuromarketing clear.