The musician Kenna grew up in Virginia Beach. His family was well-educated, and he grew up watching CNN and playing the piano. When Kenna was still a young man, he was discovered by a talent scout, who referred him to Craig Kallman, the president of Atlantic Records. Kallman had a difficult job—he had to listen to hundreds of songs a day and choose the two or three that might be hits. When Kallman heard Kenna’s music, he was convinced that Kenna would be a huge star. Kallman sent Kenna to meet with the manager of the rock group U2, among many other important musicians and producers. But there was a problem. Even though Kenna was highly popular among producers, executives, and musicians, he didn't seem to appeal to actual listeners. Kenna’s songs never “tested” well—when played for a small sample audience, the audience didn’t give him good marks. Bafflingly, Kenna’s career began to stall: even though he got glowing reviews from music professionals (i.e., people who are paid to predict what the public will like), the public itself didn’t like him. Some of Kenna’s fans—and Kenna himself—have suggested that Kenna’s career stalled because his music is difficult to categorize: it falls somewhere between indie, funk, and dance music.
In this chapter, Gladwell considers the science of polling. For the rest of the chapter, Gladwell will try to explain how test audiences, polls, and expert opinion can differ so enormously. For now, however, it’s important to pick up on the fact that Kenna’s music is very difficult to categorize—it doesn’t fall into any single musical genre. As we’ll see, test audiences often react negatively to new and radical products, ideas, and works of art—in other words, one problem with polls and test audiences is that laypeople have a bad habit of confusing “different” with “bad.”
In general, polling is an important technique. In politics, for example, politicians can use polls to appeal to their constituents. The political pollster Dick Morris tells a funny story about meeting Bill Clinton for the first time in 1977: when Clinton met Morris, he asked Morris about his polling techniques. Unusually (even for a politician), Clinton was so fascinated with the details of polling that he spent four hours talking with Morris—later, when Clinton ran for president, he hired Morris to help with his campaign. But perhaps there’s a problem with polling—as we’ve seen already, people can’t always explain why they choose to do certain things; they act intuitively. So maybe we should take polls with a grain of salt—maybe people can’t always explain what they want.
As Gladwell showed in the previous chapter, the act of explaining one’s tastes, instincts, and intuitive decisions sometimes interferes with the decisions themselves. Therefore, there seems to be an inherent problem with polls—in which laypeople are often asked to explain their intuitions about new products.
A good example of the unpredictability of polling came in the early 1980s, when Coca-Cola was trying to distinguish itself from its rival, Pepsi. In the 1970s, Pepsi had introduced the “Pepsi challenge,” in which ordinary people were blindfolded and asked to choose between small cups of Coke and Pepsi. A slim but statistically significant majority of people said they preferred Pepsi to Coke. In response to Pepsi, Coca-Cola released a new product: New Coke. New Coke “tested” very well—people were asked if they preferred New Coke or regular Coca-Cola, and they overwhelmingly claimed to prefer the former. But when New Coke was released, it sold horribly. Coca-Cola was forced to reintroduce its original product, which it called “Classic Coke.” But then, most surprisingly of all, Coca-Cola found that it had accidentally solved its biggest problem, its rivalry with Pepsi. Since the introduction of Classic Coke, Coca-Cola has remained the number-one soft drink in the world, edging out Pepsi year after year. Furthermore, ever since the introduction of Classic Coke, Coke has beaten Pepsi in the “Pepsi Challenge.” In short, the story of New Coke “is a really good illustration of how complicated it is to find out what people really think.”
“New Coke” is a notorious example of an unsuccessful product—it was one of the biggest flops in the history of business. But for Gladwell, New Coke is also an example of the pitfalls of polling and audience testing. Even after all the research indicated that the Coca-Cola Company should redesign their soda, the resulting product, New Coke, wasn’t popular at all. Test audiences claimed that they wanted New Coke, but clearly, the general public did not. As we saw with the Fismans’ research in Chapter 2, there is a fundamental gap between what people want and what they think they want. It’s the job of a successful company to navigate this gap, responding to polls and test audiences, but also taking them with a grain of salt.
There are many good examples of how the presentation of a question can change the response. For instance, the “Pepsi Challenge” was designed for participants to sip a small amount of both Coke and Pepsi. But in different versions of the test, participants were instructed to take home a case of Coke and a case of Pepsi. Studies found that people preferred a small amount of Pepsi (which is slightly sweeter than Coke) but a large amount of Coke.
Gladwell begins by poking holes in the market research that led Coca-Cola to change its recipe. Contrary to Coca-Cola executives’ beliefs, the “Pepsi Challenge” didn’t necessarily prove that people preferred Pepsi to Coke; it just suggested that, when blindfolded, people preferred a small amount of Pepsi to the same amount of Coke.
The subtleties of polling Coke and Pepsi-drinkers illustrate the concept of sensation transference, which was pioneered by the revolutionary marketer Louis Cheskin. Cheskin observed that sometimes, people “transfer” their impressions of a product’s case or packaging to the product itself. For example, Cheskin tried to find a way to market margarine—which, at the time, was a highly unpopular product, viewed as a cheap substitute for butter. Cheskin’s insight was to package margarine differently, wrapping it in gold foil and thereby making it look like butter. Cheskin’s strategy was highly successful—because his company presented margarine differently (as a fancy, shiny product), consumers unconsciously thought that margarine itself was fancy. Cheskin realized that if he had polled consumers about whether they wanted margarine to be wrapped in foil or not, they wouldn’t have admitted that they did. In short, consumers didn’t know what they wanted in margarine—Cheskin had to “tell” them by repackaging the margarine.
Louis Cheskin’s success as an advertiser is notable because it proves that people don’t always know what they want; they have to be told. In the case of margarine, for instance, people weren’t consciously aware that they would respond to margarine wrapped in attractive gold foil—it wasn’t until Cheskin released margarine of this kind that people “discovered” their attraction to such a product. It’s also important to notice that Cheskin didn’t change the taste of margarine; rather, he only altered the product’s presentation. Good packaging can be very important because it catches customers’ eyes and convinces them that the product itself is good, too.
Cheskin founded a consulting firm that still exists today. The firm has helped package hundreds of successful products. One of its most influential findings was that people prefer realistically drawn “food mascots” when they’re shopping at grocery stores. These findings helped inspire a wave of semi-realistic food mascots, such as Chef Boyardee, Orville Redenbacher, and Betty Crocker. Mascots of this kind have been instrumental in selling more canned and packaged goods in the last few decades.
Cheskin’s research into marketing and packaging has been very influential, inspiring hundreds of companies to revamp their products. Once a new package or advertisement is proven to be successful, many other companies will try to imitate that package or ad—hence the sudden popularity of semi-realistic food mascots.
It could be argued that Cheskin’s advertising techniques are dishonest. However, Gladwell argues that these techniques are no more dishonest than charging more money for a new kind of chocolate chip cookie with bigger chocolate chips. In either case, people will pay more for a product because they believe it will taste better (either because the chocolate chips are bigger or because margarine wrapped in foil seems like it might taste better). Furthermore, people decide how a product tastes based not only on its taste but also on how it looks and how it triggers memories and feelings. Therefore, it would be foolish for a company to ignore all the non-gustatory elements of food.
While it could be argued that food corporations pay millions of dollars to “con” consumers into thinking that products taste better than they really do, Gladwell argues that presentation and packaging are themselves elements of the experience of “tasting” a food. Foods don’t taste good simply because of their literal taste; they taste good because of the way they’re arranged and presented. Therefore, it’s not inherently dishonest for a company to spend a lot of money on packaging—a new package will change a consumer’s experience of the product itself.
Coca-Cola made a big mistake when it introduced New Coke—it placed too much emphasis on blind taste tests. The idea that a product is better because blindfolded people prefer it is ridiculous—nobody consumes soda blindfolded, anyway. Coca-Cola is popular not only because of its literal taste but because of the shape of the bottle, the color of the logo, the personal associations of the consumers, and the celebrities who endorse the soda. In short, Coke focused too much on the literal taste of the product and not enough on the overall Coke “brand.” Perhaps the same is true of Kenna’s career—when marketers tested Kenna’s songs, they focused too much on the songs, and not enough on marketing Kenna.
Gladwell suggests that perhaps the reason that Kenna didn’t become a bigger success is that music studios concentrated on the literal sound of Kenna’s music, and not enough on the overall experience of seeing Kenna perform. Just because test audiences preferred Pepsi to Coke didn’t mean that the public would buy more Pepsi than Coke; by the same token, the fact that test audiences didn’t like Kenna’s music doesn’t necessarily prove that Kenna couldn’t have been a big star.
In the early 1990s, the Herman Miller furniture company developed a new product: the Aeron chair. The Aeron was designed to be as comfortable as possible, with good shoulder support, adjustable arms, etc. But despite the fact that it felt comfortable, the Aeron chair didn’t look like a good chair. Consumers don’t buy chairs solely because they’re comfortable (although this is an important factor)—they tend to buy chairs that look distinguished and throne-like. The Aeron looked ugly and odd. When the chair was “tested,” consumers said that it felt comfortable, but still insisted that they hated it. Herman Miller had three choices: selling the original chair; spending a lot of money to redesign it; or not selling it at all.
The Aeron chair is an excellent example of a new, different product—altogether unlike any other product on the market. Like many new, different products, the Aeron chair “tested” poorly—test audiences had never experienced anything like it before; partly for this reason, they concluded that it was awful.
In the end, Herman Miller chose to sell the original Aeron chair, even though it had tested horribly. To the everyone’s surprise, the Aeron chair became the most popular chair in the history of the company. Even more bafflingly, it won countless design awards, and when people were asked about the chair, they described it as beautiful and elegant. The same chair that test audiences had described as hideously ugly had somehow become beautiful.
The popularity of the Aeron chair reiterates Gladwell’s central point in this chapter: there is a gap between polling and popularity. Put another way, there’s a gap between what people claim they like and what they actually like (echoing the Fismans’ research from the earlier chapter).
The Aeron chair is a good example of how market research can’t always identify what people want, because people can’t always articulate their real feelings. In testing, customers said they “hated” the Aeron chair, but perhaps that was only because they hadn’t seen an Aeron chair before—it was just “different.” Similarly, the world of television is full of stories of test marketing failures. The famous 1970s sitcom All in the Family, featuring a brash, unlikable protagonist named Archie Bunker, tested so badly with audiences that the ABC network considered dropping it. But when the show finally aired, it became a huge hit. Test audiences reacted negatively to the program simply because it was “different,” but once they were used to the show’s style of humor, they found that they enjoyed it. A major problem with test marketing is that people easily confuse “different” with “bad,” so that really revolutionary products and ideas (such as the Aeron chair) often “test” badly.
In this passage, Gladwell sums up his ideas about the flaws of polling and testing. While most polls and test audiences are accurate reflections of what the public wants and enjoys, polls and test audiences are also bad at predicting the success of really revolutionary new products. As Gladwell says here, it’s easy for laypeople to confuse a new, “different” product with a bad product. But once a new product becomes widespread and isn’t different anymore, laypeople sometimes decide that they enjoy the product after all.
Gladwell recalls meeting two professional food tasters, Gail Civille and Judy Heylmun, at a company called Sensory Spectrum. Sensory Spectrum (or SS) is a very important company: its food tasters are trained to measure precisely how different a new food is from the market norm. Gladwell took Civille and Heylmun to a French restaurant, where he was amazed by their ability to analyze tastes in great detail. When they tasted a bite of crème brûlée, for example, they described it as “missing the whole winey texture” and being “a little too woody.” Civille and Heylmun are obviously experts in food tasting—in other words, they have a sophisticated understanding of “what goes on behind the locked door of their unconscious.”
In a previous chapter Gladwell argued that, at times, it’s best to keep intuition behind a locked door—for example, we should not ask laypeople to explain their hunches and snap judgments, for fear of interfering with the processes of the adaptive unconscious. However, there are some rare kinds of people—highly trained experts—who can articulate the reasoning underlying their snap judgments. Indeed, Civille and Heylmun are paid to articulate why they do or don’t enjoy certain foods.
There was a famous psychological experiment in which food experts were asked to rank forty-four different brands of jam according to specific measures of taste and texture. Then, the scientists gave the top and bottom three jams to a group of college students and asked them to rank the jams according to the same criteria. Surprisingly, the students ranked the jams in more or less the same way as the experts—they could clearly distinguish the best from the worst. But in a second version of the test, the college students were asked to rank the same jams, and also explain the reasons for their preferences. This time, the students ranked the jams completely differently from the experts—the request that they explain their opinions turned them into “jam idiots.” The study suggests that average people are surprisingly adept at identifying what is and isn’t “good”—they just lack the sophisticated vocabulary and training to articulate why they prefer certain foods or products.
The interesting thing about this experiment is that, ordinarily, the average person is pretty good at ranking jams from best to worst (assuming, that is, that experts’ rankings are the “right” rankings). But when laypeople are asked to articulate the reasons underlying their tastes, they lose their tastes altogether. This experiment is another good illustration of the concept of verbal overshadowing that Gladwell described in the previous chapter: the rational, conscious mind interferes with the behavior of the adaptive unconscious.
The jam experiment, along with Gladwell’s interactions with Heylmun and Civille, reinforces an important point: experts and laypeople often have the same basic tastes, but while experts are good at articulating their reasons for preferring certain foods and products, average people cannot explain their preferences, and in fact, lose their ability to choose good products because they’re asked to articulate their reasons. Gladwell asks, “Isn’t this what happened to Kenna?”
Gladwell offers an important addendum to his ideas about verbal overshadowing: most of the time the rational, conscious mind interferes with people’s snap judgments. But some people—trained experts—can articulate their tastes and snap judgments. One of the basic problems with polls is that they force laypeople to behave like experts—they make people articulate why they do or don’t like something.
Gladwell returns to discussing Kenna’s career. After many years of trying, Kenna was signed by Columbia Records and released an album. The album was a modest success—the music videos were nominated for a couple awards, and the songs were sometimes played on the radio. But Kenna never managed to get his singles played on Top 40 radio, because test audiences didn’t like them. In short, the experts loved Kenna—they could explain, in very particular language, why they loved Kenna. But average people, when asked to explain their feelings about Kenna’s music, were unable to articulate their thoughts clearly, and as a result, they said that they didn’t like Kenna at all. Kenna is very popular at concerts and live performances, but he can’t win over a focus group. As a result, his career has never really taken off.
The chapter ends with a further discussion of Kenna’s career. The fact that Kenna is a charismatic guy and a great live performer suggests that Kenna really does have what it takes to become a star. But unfortunately, Kenna will never get the chance to be a star (at least at the time of the book’s writing) because his music doesn’t “test” well. It’s possible that, if music studios gave Kenna’s music a chance and played it on the radio, it would become very popular—similar to the way the Aeron chair became very popular, even though test audiences hated it.