You May Also Like

by Tom Vanderbilt

  —

  Expectations drive liking. We spend almost as much time anticipating whether we will like something as we do actually liking it. When you are told how good a movie is, two things can happen when you see it. The first is “assimilation.” This is when all those built-up expectations actually lead you to like it more than you might have otherwise. With “contrast,” on the other hand, you end up being disappointed, more than if your expectations had not been lifted to such a lofty plateau.

  With food, we tend toward assimilation. “The first taste is always with the eyes,” as the saying goes, but even before that the food has been sampled by the mind. The problem at Natick is that expectations are often so poor. In one study, Cardello’s team took Green Giant corn and put it in an MRE package, and vice versa. “People will like the corn significantly more when they think it’s Green Giant,” he says. Perhaps even more than an inherent liking for Green Giant, “that negative stereotype of military products drives the liking down.”

  Assimilation speaks to another virtual law of liking: The more a person’s experience with a product matches his expectation, the more he will like it, and vice versa. This happens all the time with food, in ways that have little to do with our actual sensory reactions to a product. Tell people a coffee is bitter, and they will think it is more bitter than if you had not told them. The opposite can happen as well, with our brains, neuroscientists have suggested, actually “suppressing” our response to the bitter when we are not told to expect it. Tell subjects that an orange juice has vodka in it, and they will like it more than the one that does not—even when neither juice actually has alcohol (I need hardly mention these were college students).

  Simply give people some kind of information about what they can expect from a food they have never eaten—in one Natick trial, Arctic cloudberries—and they will like it more. If it is “weird space food,” call it weird space food! People will still like it more (the research was done on astronauts). A Natick study had soldiers eat in the dark, a not unlikely occurrence for soldiers. They liked things more when they were told what they were eating.

  When our expectations are violated, interesting things happen. In one well-known study, people were given a salmon-flavored ice cream that was labeled simply as “ice cream” or “frozen savory mousse.” People liked it more as mousse than ice cream. Their dislike for the ice cream was so intense, in fact, that, as the researchers noted with some concern, “many participants verbally described the food as disgusting.” The idea of assimilation and contrast is why menus always announce the noticeable presence of salt in desserts with chocolate or caramel. As one prominent pastry chef noted, “If we say something is salted, it’s to call out the salt so people aren’t surprised. It gives them a chance to appreciate the contrast of salty against the sweet.” In other words, to like it more. Remember, we are primed to notice—and not like—things that are “wrong” with our food.

  But the salmon ice cream experiment shows that liking is not merely liking a thing in itself. What you like it as can be just as important. In one study Cardello conducted on “novel foods” (“The U.S. Armed Forces and N.A.S.A. frequently require the development of ‘novel’ foods for use in extreme environments”), subjects were given “soup” (Campbell’s cream of mushroom from concentrate) and a “liquid diet” (a pulverized, viscous chicken cacciatore substance developed for jaw-surgery patients—which I believe my school cafeteria also once dished up for me). Both were served in a ceramic bowl and a glass with a straw. Everything was labeled “soup.” Perhaps not surprisingly, people liked the actual soup more. But in a second trial, the substance in the glass was labeled “dental liquid diet.” Suddenly people liked the dental liquid more than the soup, when it was in the glass. As the researchers noted, “The change in expectancy, caused by the change in label, made the soup more dissonant and resulted in a reduced affective response.”

  This expectations dissonance is not limited to strange foods in military labs. One afternoon, I went to visit Garrett Oliver, the stylish, urbane, and opinionated brewmaster of Brooklyn Brewery. As we sat over a few “ghost bottles” of a one-off beer aged in a bourbon barrel and containing yeast sediment from a Riesling fermentation, Oliver told me how, a few years prior, he had come up with a new limited-run beer. The brew was based on a popular cocktail called Penicillin, which blends the flavors of scotch, ginger, honey, and lemon. “It has some sourness, some sweetness,” he said. “What I loved about it was these elements all hang together to make a harmonious whole.” He wondered if he might bottle some of that same magic in a beer. And so he blended peat-smoked malt, organic lemon juice, wildflower honey, and minced ginger.

  The response was incredibly polarized. “Draft magazine called it one of the top twenty-five beers of 2011,” he said. “But some people wanted to punch us in the head.” The problem, he suspected, was that not every bartender was presenting it in the way Oliver had imagined. “All the bartender had to tell somebody at the end of the day was that it’s based on a cocktail that has scotch-ginger-honey-lemon. They didn’t always do it.” So some consumers drank a beer that they expected to be based on a cocktail, with maybe the original cocktail even served alongside as a cue. Others simply got, as Oliver described it, “ ‘Oh, here’s a new beer from Brooklyn Brewery, probably a pale ale or something’ that did not prepare them for this weird flavor experience. They got it in their mouth and were, like, ‘bleh.’ ” They were not told how to like it.

  ON A SCALE FROM ONE TO NINE: THE PROBLEM OF MEASURING LIKING

  As we have seen, one messes with consumer expectation at one’s peril.

  One of the most notorious cases of taste and violated expectations is with Crystal Pepsi, the clear soft drink released by the beverage company in the early 1990s. The drink was inspired by rising sales of bottled waters and an identified trend toward “clear” products, ranging from dishwashing liquid to deodorant. Crystal Pepsi was positioned as a “lighter,” in both color and calorie content, more “natural” alternative to Pepsi-Cola itself. Early indications were positive. There was what the then Pepsi CEO, David Novak, described as a “hugely successful” test launch in Colorado. Three months after it was distributed nationally, Crystal Pepsi had grabbed a respectable 2.4 percent market share. It was even priced higher than Pepsi, hinting at its premium cachet.

  Then the fizz went out. By 1994, Crystal Pepsi was gone, relegated to an inglorious footnote in the history of marketing mishaps. What went wrong? Aside from the obvious—that most new products fail—there were early stirrings of discontent. A blind taste test by one newspaper hinted at the problem: People preferred the taste of Crystal Pepsi, but only when their eyes were closed. Seeing Crystal Pepsi created expectations of what it should taste like, and these expectations were clearly violated. Pepsi bottlers, Novak recalls, raised a different kind of expectation issue: Crystal Pepsi did not “taste enough like Pepsi.” The name Pepsi itself led consumers to believe it would be Pepsi-like. Calling it simply Crystal might have helped. But the episode raised a tricky question: If a food color’s “principal use,” as one study noted, is for “flavor identification,” what flavor are you identifying when you take the color away?

  Apart from the problem of violated expectations, the Crystal Pepsi debacle contains another important lesson: just how difficult it is to anticipate consumer liking. It seems a simple problem: If enough people like something in a taste test, why would they not like it in the real world? Pepsi certainly did not just chuck Crystal Pepsi into the marketplace on a whim. Some ninety people, it was said, worked fifteen months on the product, cycling through several thousand versions. And well before it even made it to a regional taste market, one can be sure it was tested in-house by any number of sensory and consumer panels, the bulk of which, one presumes, must have said that they liked it.

  As it happens, this very thing—measuring people’s liking for a product on a food company “consumer panel”—was developed and perfected at Natic
k. The program itself was founded in 1944 at Chicago’s Quartermaster Food and Container Institute, in response to an ongoing problem of ration quality and its impact on troop morale. A team of psychologists, many of whom would go on to do seminal work in the food industry, was assembled. “One of the first issues that came up,” Cardello told me, “was how do you measure how much someone likes something?”

  Pioneering psychologists like Wilhelm Wundt had tried to quantify, through “psychophysics,” the inexact ways our senses responded to various stimuli (for example, when you double the sweetness of something, and it does not taste twice as sweet).

  But no one had been able, or had much tried, to quantify liking. And so the “nine-point hedonic scale” was born. First used on soldiers, it eventually found its way into the test kitchens of just about every major food manufacturer. Whatever is in your refrigerator at this moment, chances are that someone, somewhere has indicated his liking of it on a scale from one to nine. There was, according to one account, an early attempt to introduce an “11-point scale,” but it would not fit on government-regulated paper. Humans have even been trotted out and asked to write down their scale-of-one-to-nine reaction on products like cat food. Why? Felines, as an accompanying report noted, are “clearly unable to verbalize their likes and dislikes.” They may haughtily strut away from the bowl, their tail a flag of disdain, but this gesture does not easily translate into a numerical scale. “Perhaps surprisingly,” the report concluded, “the grand mean of all hedonic scores was 4.7, placing it between the ‘neither like nor dislike’ and ‘like slightly’ scale adjectives.” People thought cat food was not too bad, at least as cat food.

  The simplicity, relative accuracy, and value of hedonic scores as an industry standard has overshadowed the ongoing methodological issues in trying to put a number on liking. Other methods, like polygraphs, have failed dismally. But issues abound. There are semantic problems. Does “like slightly” mean the same thing to one person as another? There are issues with the math. The number eight, Cardello noted, does not mean twice as much liking as the number four. Could liking and disliking be expressed on the same scale? As work by Timothy Wilson and colleagues at the University of Virginia has made clear, asking people to analyze why they chose something can lead them to change their original choice—and usually not for the better.

  But merely asking consumers what they like is also not as simple as you might think. In one common tool, the “just about right” scale, people are given samples of a product. Each will have, say, a different gradation of sweetening. The consumer indicates which is “just about right.” Sounds fine, no? There is just one problem: The level of sweetening a subject chooses is often different from what he says he likes.

  Then there is the fact that most people do not pick the number one or nine. Those seem too artificial. People hedge. It becomes, by default, a seven-point scale. “You’re never sure that you’re not going to get a product in the next sample that’s even better than one you just tasted,” Cardello said.

  Our confusion about our own tastes translates into trouble for people trying to measure those tastes. People in general tend toward a “regression to the mean” in terms of liking. Ask them ahead of time how much they like lasagna or liver, say, and then ask them again after they have actually consumed it, and subjects will mark their favorite foods a bit lower and their least favorite a bit higher. Expectations haunt our liking, but they confound us. Peer into the science of liking long enough, and you might begin to think this is something approaching a mantra: The bad is never as bad as we think it is, the good never as good.

  —

  One reason Natick has proven so influential is that, year after year, it has had an essentially captive audience of subjects to test. It is also a laboratory of pure liking, uncorrupted by the contexts of the outside world. Soldiers eating MREs do not see the price of food; they are not swayed by advertising. Nor do they have any choice. One of the research concerns has been food “monotony”—how long a soldier could be reasonably expected to eat nothing but MREs. The army’s own analysis, Darsch tells me, targeted twenty-one days. This was probably “on the conservative side,” he allowed. “You could probably go thirty-plus days and not have a statistical loss of body mass and muscle.”

  But more broadly, Natick has thought long and hard about how to plan menus that offer the most variety as is logistically feasible and that are most liked. Soldiers will not simply eat anything when they are hungry. Consumption, not to mention health and morale, drops off as food acceptability declines. Feeding an entire army means preferences must be broad and wide. As an early study observed, “Even foods that are extremely well-liked, but only by a small proportion of the consumers, are unsuited for military use.” Dishes like New England clam chowder have failed because, as Darsch put it, “a lot of the folks eating it didn’t really know what New England clam chowder was.”

  Howard Moskowitz, a prominent figure in the food industry, was working at Natick in the 1950s on mathematical models for “menu optimization.” Over breakfast at the Harvard Club in New York City, he said his inquiry was simple: “How frequently can we serve something so it doesn’t become tiring?” Menus, in his view, are driven by two dynamics: liking and time. There are things that we like, but how quickly will we tire of them? The thing most liked in a taste test, various studies have shown, often becomes the least liked after a number of samples. Crystal Pepsi might have seemed fresh and interesting in a taste test, but was it actually something consumers would restock the fridge with? That intense sugar rush or novel flavor may seem great the first time, “but you have to live with it,” Moskowitz said.

  “If you like something a lot more,” he continued, “do you choose it more often?” Not necessarily. We begin to pick things we may like much less, perhaps as a way of protecting our liking for that loved thing. One wants to avoid “death by hamburger,” as he dubbed it. Why should we even tire of a particular food? I asked. Is it, per sensory-specific satiety, that our nutritional needs are being met? Is there some innate desire for novelty? “I don’t know,” he said, sighing. “Why do we habituate to the smell of a fragrance? Why is it when we sit in a house next to the railroad tracks, we don’t hear the railroad anymore?” Why should we need choice? “Go to a diner,” he said. “Diners have menus with seven pages. But you order the same thing. You don’t want choice. You want the illusion of choice.”

  Tyler Cowen, perhaps our most food-aware economist, noted that he is often puzzled to hear, as that day’s lunch choice is being deliberated, someone utter something like “I don’t want Thai food today; I just had that yesterday.” This rather neglects the fact that Thais are eating Thai food every day. “Would it be so terrible,” he wondered, “to eat only Indian food, whether at home or in restaurants, every day for a week?” Often, when we think we are tired of something, we may simply be forgetting how much variety we have actually had (in a phenomenon that has been called “variety amnesia”). Curiously, while one might expect people to tire less quickly of flavorful food, food monotony research at Natick has shown the opposite: The more bland it is, the less quickly soldiers grew tired of it. Bland food, after all, fades from memory more quickly than exciting food. The less you remember having it, the less tired of it you get.

  Natick also had to grapple with where food was being consumed. The very same food will be rated higher when served in a restaurant versus an institutional cafeteria or a lab. Soldiers in the field face two challenges: Not only are they eating MREs, with their limited variety and entrées of approximate flavor and dubious texture, but they are often eating them in the far-flung, inhospitable environments for which they were designed. In a series of groundbreaking experiments, a group of soldiers (bivouacked on an island in Hawaii) and MIT students (on campus) ate nothing but MREs. The soldiers ate them for thirty-four days straight, the students for forty-five days. Both groups deemed the food “acceptable” (which did not speak well for MIT’s canteen). Both groups l
ost weight. The students, however, ate more than the soldiers in the field. The experiments showed the importance of context on liking. For many reasons, it is more difficult to get people in the field to eat.

  Context is no less important in the real world. People eating in an ethnic restaurant with appropriate decor rate the food higher; add some red-checkered tablecloths or a Sergio Leone poster, they eat more pasta. The loudness, and type, of music can affect the way we feel about our food. We eat more when we are in larger groups. The type of glassware, the weight of plates, whether the color of the food matches the color of the plate—even how long people have to wait for their meal—all have been shown to influence how much we like, and eat, food.

  There is a poignant scene in the film Sideways in which Miles, the hapless protagonist, in a fit of pique and despair over his dismal life prospects, brings his treasured bottle of 1961 Cheval Blanc to a fast-food joint. Amid the harsh light and the smell of grease, to the accompaniment of a burger and onion rings, he surreptitiously quaffs his “special occasion wine” from a Styrofoam cup. The wine is still the same wine, and if consumption were always just about the thing being consumed, the level of enjoyment should theoretically be the same. But all the context factors are “off”: He is alone, he is eating mediocre food, he does not have a proper glass, the decor is terrible. He is drinking with vengeance, not appreciation.

  Context is not just place but time. Your love of breakfast cereals probably does not, in normal circumstances, extend to dinner. Breakfast itself is a rather strange meal, as the Dutch researcher E. P. Köster has observed. The most adventurous gourmands will eat the same thing for breakfast, day after day. They would hardly contemplate this at dinner. Sheer convenience explains much of it, to be sure, but research suggests there are whole classes of textures that are less liked at breakfast, varying by culture. By the time our after-dinner dessert rolls around, we are hungry for variety. It is as if we wake up less primed to desire novelty, our threshold for excitement slowly ramping up as the day progresses.