This time the reception was downright chilly. “They didn’t understand the aesthetic at all,” says Dowell. Herman Miller was told to cover the Aeron with a solid fabric and that it would be impossible to sell it to corporate clients. One facility manager likened the chair to lawn furniture or old-fashioned car-seat covers.

Another said it looked as though it came from the set of RoboCop, and another said that it looked as if it had been made entirely from recycled materials. “I remember one professor at Stanford who confirmed the concept and its function but said he wanted to be invited back when we got to an ‘aesthetically refined prototype,’” Dowell remembers. “We were behind the glass saying, ‘There isn’t going to be an aesthetically refined prototype!’”

Put yourself, for a moment, in Herman Miller’s shoes. You have committed yourself to a brand-new product. You have spent an enormous amount of money retooling your furniture factory, and still more making sure that, say, the mesh on the Aeron doesn’t pinch the behinds of people who sit in it. But now you find out that people don’t like the mesh. In fact, they think the whole chair is ugly, and if there is one thing you know from years and years in the business, it is that people don’t buy chairs they think are ugly. So what do you do? You could scrap the chair entirely. You could go back and cover it in a nice familiar layer of foam. Or you could trust your instincts and dive ahead.

Herman Miller took the third course. They went ahead, and what happened? In the beginning, not much. The Aeron, after all, was ugly. Before long, however, the chair started to attract the attention of some of the very cutting-edge elements of the design community. It won a design of the decade award from the Industrial Designers Society of America. In California and New York, in the advertising world and in Silicon Valley, it became a kind of cult object that matched the stripped-down aesthetic of the new economy. It began to appear in films and television commercials, and from there its profile built and grew and blossomed. By the end of the 1990s, sales were growing 50 to 70 percent annually, and the people at Herman Miller suddenly realized that what they had on their hands was the best-selling chair in the history of the company. Before long, there was no office chair as widely imitated as the Aeron. Everyone wanted to make a chair that looked like the exoskeleton of a giant prehistoric insect. And what are the aesthetic scores today? The Aeron is now an 8. What once was ugly has become beautiful.

In the case of a blind sip test, first impressions don’t work because colas aren’t supposed to be sipped blind. The blind sip test is the wrong context for thin-slicing Coke. With the Aeron, the effort to collect consumers’ first impressions failed for a slightly different reason: the people reporting their first impressions misinterpreted their own feelings. They said they hated it. But what they really meant was that the chair was so new and unusual that they weren’t used to it. This isn’t true of everything we call ugly. The Edsel, the Ford Motor Company’s famous flop from the 1950s, failed because people thought it looked funny. But two or three years later, every other car maker didn’t suddenly start making cars that looked like the Edsel, the way everyone starting copying the Aeron. The Edsel started out ugly, and it’s still ugly. By the same token, there are movies that people hate when they see them for the first time, and they still hate them two or three years later. A bad movie is always a bad movie. The problem is that buried among the things that we hate is a class of products that are in that category only because they are weird. They make us nervous. They are sufficiently different that it takes us some time to understand that we actually like them.

“When you are in the product development world, you become immersed in your own stuff, and it’s hard to keep in mind the fact that the customers you go out and see spend very little time with your product,” says Dowell. “They know the experience of it then and there. But they don’t have any history with it, and it’s hard for them to imagine a future with it, especially if it’s something very different. That was the thing with the Aeron chair. Office chairs in people’s minds had a certain aesthetic. They were cushioned and upholstered. The Aeron chair of course isn’t. It looked different. There was nothing familiar about it. Maybe the word ‘ugly’ was just a proxy for ‘different.’”

The problem with market research is that often it is simply too blunt an instrument to pick up this distinction between the bad and the merely different. In the late 1960s, the screenwriter Norman Lear produced a television sitcom pilot for a show called All in the Family. It was a radical departure from the kind of fare then on television: it was edgy and political, and it tackled social issues that the television of the day avoided. Lear took it to ABC. They had it market-tested before four hundred carefully selected viewers at a theater in Hollywood. Viewers filled out questionnaires and turned a dial marked “very dull,” “dull,” “fair,” “good,” and “very good” as they watched the show, with their responses then translated into a score between 1 and 100. For a drama, a good score was one in the high 60s. For a comedy, the mid-70s. All in the Family scored in the low 40s. ABC said no. Lear took the show to CBS. They ran it through their own market research protocol, called the Program Analyzer, which required audiences to push red and green buttons, recording their impressions of the shows they were watching. The results were unimpressive. The recommendation of the research department was that Archie Bunker be rewritten as a soft-spoken and nurturing father. CBS didn’t even bother promoting All in the Family before its first season. What was the point? The only reason it made it to the air at all was that the president of the company, Robert Wood, and the head of programming, Fred Silverman, happened to like it, and the network was so dominant at that point that it felt that it could afford to take a risk on the show.

That same year, CBS was also considering a new comedy show starring Mary Tyler Moore. It, too, was a departure for television. The main character, Mary Richards, was a young, single woman who was interested not in starting a family—as practically every previous television heroine had been—but in advancing her career. CBS ran the first show through the Program Analyzer. The results were devastating. Mary was a “loser.” Her neighbor Rhoda Morgenstern was “too abrasive,” and another of the major female characters on the show, Phyllis Lindstrom, was seen as “not believable.” The only reason The Mary Tyler Moore Show survived was that by the time CBS tested it, it was already scheduled for broadcast. “Had The MTM been a mere pilot, such overwhelmingly negative comments would have buried it,” Sally Bedell [Smith] writes in her biography of Silverman, Up the Tube.

All in the Family and The Mary Tyler Moore Show, in other words, were the television equivalents of the Aeron chair. Viewers said they hated them. But, as quickly became clear when these sitcoms became two of the most successful programs in television history, viewers didn’t actually hate them. They were just shocked by them. And all of the ballyhooed techniques used by the armies of market researchers at CBS utterly failed to distinguish between these two very different emotions.

Market research isn’t always wrong, of course. If All in the Family had been more traditional—and if the Aeron had been just a minor variation on the chair that came before it—the act of measuring consumer reactions would not have been nearly as difficult. But testing products or ideas that are truly revolutionary is another matter, and the most successful companies are those that understand that in those cases, the first impressions of their consumers need interpretation. We like market research because it provides certainty—a score, a prediction; if someone asks us why we made the decision we did, we can point to a number. But the truth is that for the most important decisions, there can be no certainty. Kenna did badly when he was subjected to market research. But so what? His music was new and different, and it is the new and different that is always most vulnerable to market research.