-By Jim Edwards



The human brain is smaller than most people think. If you were to peel away the hair, skin and muscle fiber, then remove the skull, you'd be left with a mass of tissue slightly larger than a man's fist. I happen to know this because Prof. Joy Hirsch has a plastic model of one sitting in front of me, next to a can of Diet Coke, on her desk at Columbia University Medical Center.

Brains are jellylike and gray. Confusingly, Hirsch's model is hard and pink, but it'll have to do. Hirsch, the director of Columbia's Functional Magnetic Resonance Imaging Research Center, is using her plastic model to explain to me what's going to happen in a few minutes from now, when she'll insert my head into one of her three fMRI scanners down the hall.

I have come to Hirsch's book-cluttered office, located in the basement of the hospital's Neurological Institute Building in upper Manhattan, because I have a question: "Does my brain handle the brands I like differently from the brands that I do not like?" Hirsch has custom-designed a small demonstration to provide an answer for Brandweek's readers.

On one level, the answer is trivial: Of course it does. That's why I know that I like some brands and dislike others in the first place.

But if Hirsch's fMRI showed that my brain uses different structures for different brands, it could mean problems for a company that wants consumers to value its brand highly. Changing a consumer's mind-persuading him to admire a brand that he previously held in low esteem-might involve producing complex physical or biological changes inside the brain.

Of course, the desire to see inside the working mind of a consumer is something marketers have been dreaming about for many years. What's significant is that technology has finally caught up to the aspirations and made it all possible-or at least possible to begin. Thirty years ago, for instance, researchers would measure pupil dilation and constriction as consumers watched ads, or hook them up to electrodes that detected skin dampness, in the hope that their bodies would give away their level of excitement. But the results of these early stabs at neurologically centered marketing were too primitive to be of much use. That's why analog tools like the focus group and the opinion poll have dominated market research. But the fMRI could well change all that.

More Than a Feeling

Marketers have traditionally assumed that we all have an "emotional" attachment to some brands that trumps our rational shopping choices. This is why Apple users happily pay hundreds of dollars more for a computer that does pretty much exactly the same things that a Dell computer does. Conversely, marketers believe brands we dislike are relegated to background noise. Apple people ignore Dell because, well, only squares buy from Dell, right? There's even a name for that noise: "clutter."

The problem with this theory is that's exactly what it is: theory. It's a presumption that needs to be tested. Hirsch's scanner is going to do that by chopping my brain horizontally into two-dozen slices, each one 4 mm thick. The machine will then dice my brain further, into "voxels," or units about the size of a rice grain. When I (which is to say, my brain) am exposed to brands that I like the most and ones I like the least, some of those voxels will call for extra blood. In this manner, the imager will be able to see which bits of my brain are active and which are simply twiddling their neural thumbs.

Though Hirsch's work marks the inaugural stages of this type of study, its marketplace applications are not lost on her. "We're opening the doors to do commercial work," she said. "It's extraordinarily exciting."

Hirsch, who is a neuroscientist/psychologist with 10 years experience in functional imaging, isn't the only researcher who's interested in the nascent field of "neuromarketing." Nor is fMRI the only tool for conducting research in it. In Tenafly, N.J., Lee Weinblatt's PreTesting agency records high-frequency eye vibration to gauge consumers' reaction to advertisements. In Westport, Conn., EmSense uses EEG, heart rate, temperature and eye-blink monitoring to do the same thing. NeuroFocus of Berkeley, Calif., uses all of the above, including "dense array" EEG, which it believes is better.

The one thing they all have in common is a belief that, if proven to be true, could well upturn much of conventional marketing research as we know it: That consumers' non-conscious responses to marketing may be more reliable indicators of brand preference than any number of opinion polls or focus groups you'd care to convene.

Welcome to the Machine

Most of Hirsch's research is medical. She once examined a man who had "alexia without agraphia"-a stroke left him unable to read. The cruelty of the condition was that, because the human brain handles reading and writing within separate structures, the man could still write with perfect clarity-even though everything he left on the paper appeared to him as indecipherable scrawl. Prior to his stroke, the man had run a newsstand in Times Square. After it, he could not understand the titles of the magazines he sold. With the help of Hirsch's colleagues, he was able to memorize the location of every periodical, and return to work.

As Hirsch related this story, she grabbed the plastic brain model off her desk and separated it into two halves to demonstrate what the fMRI machine would soon be doing. When she finished, rather than slotting the lobes back together, Hirsch left them lying loose on the desk amid her paperwork and the Diet Coke can. It's a slightly disconcerting image-especially when it's your brain that's about to be sliced by a machine into grains of rice.

Hirsch's scanners are giant circular magnets whose enormous weight and powerful force fields have banished them to the windowless complex of small rooms and narrow corridors that snake beneath the neurological institute. It's a sterile, institutional setting where the only light comes from recessed ceiling fixtures; you wouldn't want to be left alone down here if the power went out. The fMRI control room is filled with wires, cables and computer equipment, all of it piled on top of one another, garage-sale style. Powering the works are several towers of Sun Microsystems processors, which live on their own inside a separate, icily air-conditioned room.

The electromagnetic field created by the machine can cause anything made of metal to become a projectile. In 2001, a 6-year-old boy undergoing an MRI at Westchester Medical Center died when an oxygen tank accidentally left inside the room flew into his skull, shattering it. So, I'm admonished to remove all metal objects from my pockets-keys, cell phone, spare change-and leave them on Hirsch's desk, next to her plastic brain.

An attendant then helps me onto a narrow bed and thoughtfully tapes my head between two boards to keep it perfectly still. Immobilized, I stare upwards as a plastic cage is placed over my face. Once the assistant leaves the room, the platform slides into the center of the fMRI machine, which I cannot be the first to liken in appearance to an enormous bagel. The space is claustrophobic but tolerable. The magnet takes snapshots of my head in a series of loud electronic buzzes and bangs, as a liquid helium pump to cool the machinery retorts with shush-shush-shush.

Hirsch's staff shows me a series of 40 images of brands projected onto a screen at my feet. I can see them through a mirror inside the fMRI, much as if I were peering through a periscope. I have designated the brands in advance of the test either "high value" or "low value." Some of my appraisals were obvious; Tiffany fared better than Zales. Some are more subjective. I happen to like Genesee Cream Ale and dislike Bud Light; most men, I wager, would have it the other way around. Hirsch has allowed me to furnish some that are unique to me. Liverpool F.C., my hometown team, goes up against Manchester United, our despised rivals. (What if the fMRI shows some traitorous part of my brain responding positively to Man. U.? Could it be surgically removed?)

Images of the brands click past in blocks-five high-value followed by five low, with rest periods between to establish a baseline. While I'm looking at the brand images-logos and product shots-Hirsch has asked me to rate each brand on a 1 - 10 scale to ensure that my brain is doing the same thing in each exposure, and to confirm which are high and which are low. All the while, the fMRI is slicing, slicing, slicing, looking for voxels that want extra blood.

Dissenting Votes

This type of research has its critics. Among their arguments is the point that lying supine inside a 10-ton magnetic imager is hardly a workaday experience-it's nothing like shopping at a supermarket.

"The environment of the fMRI can generate emotions in itself," offered Prof. Baba Shiv of the Stanford Graduate School of Business. He remains a realist about prospects for neuromarketing. One of his best-known experiments involves giving sips of wine to consumers lying inside an fMRI. Some of the subjects are told the wine is expensive, others that it's cheap-but they're all getting the same wine. Unsurprisingly, the brain's pleasure centers are activated to a greater extent when people believe that the wine they're sipping is pricey.

As word of Shiv's experiment got out, he started getting calls from vintners in Napa, who were excited. "Perhaps my wine gives more pleasure, and I can charge more!" he recalled them saying. And so Shiv developed a sobering talk to make the limitations of his neurological work clear to marketers who get too excited. "You need to have repeated measures with small variations in what's presented," he cautioned, and each step needs to be validated and statistically robust. That takes a lot of time and even more money. "And that's when they realize it's not a magic bullet," Shiv said. "You're not going to get an answer." (Besides, he points out, drinkers don't even taste wine when they're purchasing it-they're standing in a liquor store looking at the labels.)

Real-life brand-purchasing decisions are skewed and colored by countless ancillary stimuli, from a ringing cell phone in one's pocket to a crying baby in the mall food court. It's rare that any shopper is able to concentrate on a brand with the mental focus I achieve inside the machine.

That's why Leo Weinblatt, CEO of PreTesting, prefers measuring "saccadic" eye movement. Our eyes don't just point like binoculars at the things we want to see. Rather, they dodge and dart at high frequency, scanning all around an image. By monitoring saccade levels, Weinblatt claims he can tell which parts of a commercial a consumer is actually looking at, which are being ignored, and whether attention levels are rising or falling. The strongest argument for Weinblatt's method is also the strongest against Hirsch's: Subjects whose eye movements are being studied need not be immobilized inside a loud, electric bagel. They just watch TV using a remote to change channels.

But Weinblatt's method isn't perfect, either. It's best used for TV and that's just fine, except that many marketers are devoting an ever-increasing portion of their ad budgets to newer media: the Web, of course, and mobile phones, too.

That's where EmSense steps in. The company employs a headband embedded with sensors that record physiological stats such as EEG, heart rate, body temperature and even the rate of eye blinking, all as the participant interacts with the brand being tested."It brings us out of the research facility into the natural habitat," said chief analytics officer Elissa Moses. "It can also be camouflaged easily in a hat because it's so small. No one has to notice it."

Critics have noticed it, and they argue the device's usefulness is not unlike employing a polygraph test to determine a subject's truthfulness. An increased heart rate and sweaty palms are a sign of something-but is that something necessarily a lie? When it came to the EmSense test, Joe Kades initially counted himself among those skeptics, although now he's an EmSense customer. "Most of my career was in the ad world, and I never really had a high opinion of quantitative testing," said Kades, Virgin Mobile's vp of strategic planning and consumer insights. "I'd never seen a case where diagnostics had ever improved a campaign."

EmSense is often mentioned in the same breath as NeuroFocus, whose signature device features 64 EEG censors that cover the wearer's head "like grandma's swim cap," according to CEO A.K. Pradeep. (Full disclosure: NeuroFocus is owned by Nielsen, the parent company of Brandweek.) The device also monitors saccades and skin response. Pradeep claims he can track not only a user's attention span, but also his emotional engagement and memory.

Despite their technical differences these companies are all in accord on one point: The direct measurement of brain activity, they claim, is more accurate than consumers' rationalizations of their own behavior. And they've managed to get major brands to listen. While EmSense will not disclose its other clients, it said that they include marketers of food, beverage telecommunications, packaged goods and personal care products. NeuroFocus counts ESPN and "Fortune 100 companies across the consumer packaged goods, food and beverage, financial services, automotive and retail sectors" among its clients, which it would not identify further.

The Results

A couple of days later, Hirsch has my results. She has filtered them to show only unique brain activity triggered by the high-value brands, and the same for the low-value brands. Activated areas glow red and white; inactive ones appear green and blue. The slices, with their spectral dapples here and there, "are really beautiful," Hirsch said.

There are caveats: This was a test performed with one subject-me-so it obviously proves nothing about consumer brains in general. Nonetheless, the test was performed under real conditions and the results raise valid questions, Hirsch says. The most striking part: My brain processed high-value brands on its left side, handling the low-value ones on the right. That's not what one would expect, since the so-called left-brain is traditionally associated with conceptual processing and the right with emotions.

In my case, the high-value brands activated three areas: my left angular gyrus, left dorsolateral prefrontal cortex and my left orbitofrontal gyrus. Those systems are associated with the extraction of meaning, conceptual organization, and reward, respectively. "I haven't seen that before," Hirsch said. "My curiosity is piqued by that."

Mine, too. When it came to brands I dislike, I apparently really disliked them. My right insula-an internal fold just over an inch in length-is lit up like a runway. "Now, your brain isn't that big, but it's devoted on your right side to something where you're saying, 'That's a low-value item for me,'" Hirsch explained. Worse, the insula is understood to handle feelings of disgust. "That is not a result to be trifled with," Hirsch said.

And that's the big surprise: These results are the exact opposite of the received marketing wisdom. I'm not, apparently, emotional about brands I like. Instead, my brain behaves like an antiques dealer sifting an estate sale for high-priced items. My emotional feelings-specifically, disgust-are reserved for the brands I dislike. And I don't merely ignore those brands like clutter; I process them through the area of my brain that helps me avoid rotten food and poisonous berries.

A Broader Relevance

What do mere pockets of exaggerated activity in an organ as poorly understood as the human brain really say about an activity like shopping or, by extension, by marketing that's supposed to encourage it?

The counter-intuitive nature of the results hints at the existence of a dark universe of intimate consumer information that marketers have so far only guessed at. Suppose that our emotions aren't used to find brands we love but avoid brands we dislike. Then all those companies that have sunk millions into developing "lovemarks" and "touchpoints" may have wasted their money. Indeed, they might have done better to encourage consumers to find "hatemarks" and "loathepoints" on competing brands instead.

We don't know this from a single test, of course. But it illustrates the promise of neuroscience: Marketers need no longer be guided by supposition. They have a clearer view into the brain. That's a significant step given how frequently consumers seem to make previously "unthinkable" decisions.

For instance, back on March 3 of this year, it was widely believed that black Americans would vote largely for Hillary Clinton in the Democratic primaries because of the former First Lady and her husband's historic support in that demographic. Then on March 4, Barack Obama won in Ohio, and that "historic support" evaporated like Scotch mist.

Hirsch believes the results tell us something about "neural value," or how the brain handles concepts to which we have attached an abstract importance (which, basically, is the definition of the word "brand").

Clearly, the results of my fMRI session raise questions for marketers. How hard would it be for Anheuser-Busch to move Bud Light from my disgusted insula into my reward-loving left oribitofrontal gyrus? Could a client create incentives for agencies who can show that their brands are shifting in consumers' minds from one area to another in response to their marketing?

"We could study various approaches to changing the neurocircuitry associated with 'low value' judgments to 'high value' judgments," offered Hirsch. "I like this idea."

All it needs are companies willing to support such research, Hirsch added. (In fact, Hirsch's office already has interest from Wall Street firms who want to know why traders who bet on a losing stock so often pile on to their declining position in an attempt to prove that they were right.)

But the big takeaway is how unexpected the results were. It's common knowledge that different brain structures have discrete functions, but seeing those brands segregated from each other on different hemispheres-in a neural apartheid-was a shock. I felt guilty for not shopping at Zales. (Man. U., not so much.)

Hirsch seemed pleased at my reaction. "This is why we have science. Our intuitions about how the brain works are often not consistent with the objective evidence." The plastic model of a brain, which was lying in pieces when I last saw it, has been restored to its pedestal at the side of her office.

-------------------

Sidebar: The Case for Subliminal Brand Awareness

When it comes to studying how our brains interpret branding messages, fMRI machines represent up-to-the-minute technology. But some of the most interesting work in the field is being done in decidedly low-tech fashion by Prof. Gavan Fitzsimons of Duke University's Fuqua School of Business, Durham, N.C.

Fitzsimons' theory is that the majority of our exposure to brands occurs outside our conscious awareness. The average American is exposed to at least 3,000 brands a day-either indirectly or via advertising. That is, obviously, far more than we can rationally acknowledge. But do these peripheral exposures actually affect our behavior? Fitzsimons believes they do.

His proof comes from an experiment he and his team designed in which subjects take a peripheral-vision test. On a computer screen, boxes randomly flash in the corners, left and right, top and bottom. The subjects must call out the location of the box when they see it. At the same time, subjects are asked to keep track of a stream of numbers that scrolls in mid-screen.

What the test's participants don't know is that they are also seeing extremely brief flashes of a brand logo-either Apple's or that of IBM. The flashes last less than 30 milliseconds-faster than the brain can consciously register, but slow enough to be detected by the eye. In other words, the flashes are subliminal.

Shortly afterwards, the subjects complete what's termed an "unusual uses task." Fitzsimons asks them to think up as many nontraditional uses for a household brick as they can. "Paperweight, doorstop, break through a locked window" are the most common replies. (Fitzsimons' personal favorite is the subject who said, "Tie to the foot of my boss and sink him to the bottom of a lake.")

Then things get interesting. Fitzsimons has found that study participants who were subliminally exposed to the Apple logo generate answers 30-35% more creative than those who were exposed to the IBM logo. The relative measure of creativity is determined by a separate team of judges who know nothing of the test. (These results, incidentally, are consistent even with users who express a preference for PCs.) Fitzsimons believes that something to do with Apple's strong association with creativity motivates people to literally "think different."

To eliminate the possibility that his findings might be isolated, Fitzsimons has repeated the experiment using the Disney logo facing off with the logo for the E! cable network. Instead of creativity, subjects are tested on their honesty, because Disney, for many Americans, is associated with honesty and E!, well, less so. After the bogus visual task, subjects are asked to agree or disagree with statements such as, "I find it easy to get along with obnoxious people," or "I've never pretended I was sick to get out of doing something." Give the "wrong" answer-i.e. agree with either of those statements-and you must be lying.

According to Fitzsimons' evaluation, subjects who were exposed to Disney answer 15% more honestly than those who were flashed the logo for the channel E!.

A number of brands have taken note of Fitzsimons' experiments. "I have spoken with folks at Procter & Gamble about some of the work they've been doing on short brand exposure," he said. "They're very interested in the implications."

There's only one drawback: His research indicates that brands affect behavior, but not necessarily buying behavior. That appears to be influenced by a panoply of factors extraneous to those in Fitzsimons' study. For instance, he said, "People are impacted by price." Except, since Fitzsimons' work began, his team members have switched their computer gear from PCs to Apple. "I'm now an official Apple person," he said. "I have a laptop, desktop and an iPhone. It can't hurt, I figured."

Labels: advertisement, Music, Sociology