by Michael Moss
Howard Moskowitz had already been working on projects for General Foods for a number of years, helping the company develop winning formulas for its cereals and Jell-O, when the company called on him in 1986 to help with a more pressing crisis. Maxwell House, their flagship coffee brand, was losing badly to Folgers, and the coffee managers were at a loss about how to turn the tide. The problem was not marketing. It was far worse than that. A string of taste tests showed that people simply liked Folgers better. Pressed by their new bosses at Philip Morris, the General Food executives knew there was only one way out: They needed a new formula. Whatever beans and roasting process the company was using, it wasn’t working. They needed to start over.
Instead of making a few different roasts and submitting them to a new panel of tasters, Moskowitz pored over the data from the tests that had been done. In these, and in subsequent tests, he made a key observation. The data showed that people had varied preferences for coffee that could be grouped into three different roasts, weak, medium, and strong. Each roast was considered equally perfect by their respective fans. This was a novel concept at the time. The American consumer was viewed as a singular target, uncomplicated by variation, and every food company making every grocery product was focused on finding the one perfect formulation. Moskowitz, in a bold stroke, convinced General Foods that it should be selling not one but all three of these roasts—a breakthrough that the executive in charge of fixing Maxwell House at the time, John Ruff, told me saved the brand. “We actually reversed a loss to a win against Folgers,” he said.
If coffee had not one but three states of perfection, Moskowitz asked, what about the rest of the grocery store? Couldn’t the same principle apply there? He wasn’t envisioning the line extensions that companies later adopted to boost sales, using slight variations in color or taste or packaging to bring new consumer excitement to the main product; he imagined reworking the main products themselves, with the idea that consumers could be sorted into groups with distinct preferences. With that insight, Moskowitz’s shop turned into the industry’s miracle maker as food companies ditched their own in-house food technicians to retain his advice. Vlasic, the pickle maker, hired Moskowitz and came away with the finding that pickle lovers fell into three large groups whose preference for tartness ranged from weak to strong. Campbell, the soup manufacturer, brought him aboard to revamp its Prego spaghetti sauce, which was getting trounced by Ragu.
The brilliance of his work on Prego was memorialized in a 2004 presentation by the author Malcolm Gladwell at the TED Conference in Monterey, California, in which Gladwell called Moskowitz a “personal hero”:
After … months and months, he had a mountain of data about how the American people feel about spaghetti sauce.… Did he look for the most popular brand, variety of spaghetti sauce? No,… instead he looked at the data and he said, “Let’s see if we can group all these different data points into clusters. Let’s see if they congregate around certain ideas.” And sure enough, if you sit down and you analyze all this data on spaghetti sauce, you realize that all Americans fall into one of three groups. There are people who like their spaghetti sauce plain. There are people who like their spaghetti sauce spicy. And there are people who like it extra chunky. And of those three facts, the third one was the most significant, because at the time, in the early 1980s, if you went to a supermarket, you would not find extra chunky spaghetti sauce. And Prego turned to Howard, and they said, “Are you telling me that one third of Americans crave extra chunky spaghetti sauce, and yet no one is servicing their needs?” And he said, “Yes.” And Prego then went back and completely reformulated their spaghetti sauce and came out with a line of extra chunky that immediately and completely took over the spaghetti sauce business in this country. And over the next ten years, they made $600 million off their line of extra chunky sauces. And everyone else in the industry looked at what Howard had done, and they said, “Oh my god, we’ve been thinking all wrong.” And that’s when you started to get seven different kinds of vinegar and fourteen different kinds of mustard and seventy-one different kinds of olive oil.
And then eventually even Ragu hired Howard, … and today [with Ragu] there are thirty-six in six varieties. Cheese. Light. Robusto. Rich and Hearty. Old World Traditional. Extra Chunky Garden. That’s Howard’s doing. That is Howard’s gift to the American people.… He fundamentally changed the way the food industry thinks about making you happy.
Well, yes, and no. One thing Gladwell didn’t mention is that the food industry already knew some things about making people happy—namely, sugar. The Prego sauces—whether cheesy, chunky, or light—have one feature in common: The largest ingredient, after tomatoes, is sugar. A mere half cup of Prego Traditional, for instance, has more than two teaspoons of sugar, as much as three Oreo cookies, a tube of Go-Gurt, or some of the Pepperidge Farm Apple Turnovers that Campbell also makes. It also delivers one-third of the salt recommended for a majority of American adults for an entire day. Some of the meat versions of Prego have even higher amounts of sugar and salt, along with nearly half a day’s recommended limit for saturated fat. In making these sauces, Campbell supplied the ingredients, including the salt, sugar, and fat, while Moskowitz supplied the optimization technique and his deep knowledge of sugar. “More is not necessarily better,” Moskowitz wrote in his own account of the Prego project. “As the sensory intensity (say, of sweetness) increases, consumers first say that they like the product more, but eventually, with a middle level of sweetness, consumers like the product the most (this is their optimum, or ‘bliss,’ point).”
In the food industry, finding the bliss point for sugar in dinner products like pasta sauce would soon become passé. Products for meals were relatively easy. People had to eat dinner, so pasta sauces merely had to be more exciting than the rival brand. Snack products, by contrast, were a much bigger challenge. They were expendable—in theory, at least—and therefore required the most powerful sensory drivers. As snacks moved toward the $90 billion market they are today and the pressure for profit rose on everything else in the grocery store, food manufacturers sought formulations that would do more than make people happy. They wanted formulations that would increase desire.
It was on this front that Howard Moskowitz made his most lasting mark, starting with an investigation he undertook in 2001 to determine the factors that drove people not merely to like their food but to eagerly snatch it up. The research was funded by the ingredient giant McCormick, and Moskowitz gave it a title that reflected the industry’s drive to make people ecstatic about their food: “Crave It!”
He conducted the study with a New Jersey–based food development expert, Jacquelyn Beckley, and together they sought to identify exactly what it is about certain foods that takes us to this level of desire. They gathered consumer views on cheesecake, ice cream, chips, hamburgers, and pretzels—some thirty grocery icons in all. The resulting reams of data turned up results that not only serve as a guide for food manufacturers who want to know why cinnamon buns are so alluring; they also shed light on the very underpinnings of the obesity crisis. Because what Moskowitz found is that hunger is a poor driver of cravings. We rarely get in the situation where our body and brain are depleted of nutrients and are actually in need of replenishment. Rather, he discovered, we are driven to eat by other forces in our lives. Some of these are emotional needs, while others reflect the pillars of processed food: first and foremost taste, followed by aroma, appearance, and texture.
As disparate as these pillars may seem, one ingredient—sugar—can do it all.
Howard Moskowitz slid into a booth in a diner near his office in White Plains, where we had adjourned for lunch. We were joined by Michele Reisner, his vice president for research. The waitress suggested the Reuben sandwich, but all three of us thought better of that. I settled for the turkey club. Reisner ordered an egg-white omelet with multigrain toast. Moskowitz, who said he was watching his weight, asked for a plate of turkey breast with gravy on the side. I asked
him for the particulars of his diet. “I try not to eat potatoes,” he said. “I eat bread, but not too much. I try to eat healthfully. We have diabetes in the family.”
I ordered three cans of Dr Pepper for the table, not wanting to miss the chance for a tasting with the man who had reversed the company’s fortunes. But Moskowitz demurred. “I’m not a soda drinker,” he said. “It’s not good for your teeth.” The waitress was on my side, however, and she brought out the regular Dr Pepper along with a brand-new flavor, Dr Pepper Cherry. Relenting, Moskowitz sipped some of each, grimaced, and searched his brain to explain the trouble that his taste buds were having. “I find it terrible, really,” he said. “The cherry is overwhelming. A lot of stuff in there. Like something took.… Just awful.” He was trying to describe what was making him so unhappy.
“Benzaldehyde,” he said a few moments later. “It’s common benzaldehyde, which gives it an almond and cherry flavor. This is not in the same class as Coke.” Reisner took a few sips too and confessed that she only likes Coke, and the diet version at that. When I asked her what she thought of the taste, she simply shrugged.
Back in their office, after lunch, Reisner was blasé about their not being members of the Pepper cult. In fact, she said, the soda maker fully understood that the unique flavor of Dr Pepper was not to everyone’s liking—if it was, it would be selling as well as Coke. It has a niche that it’s trying to grow by degrees, or at least maintain, which is what Moskowitz was told when Cadbury asked for his help with Dr Pepper back in 2004. Its primary goal was not stealing customers from Pepsi or Coke. Rather, Cadbury wanted a flavor that would coax existing Dr Pepper fans into trying something new—and, with any luck, expand the reach of the brand. “This was basically, ‘We’ve got our users, and we want to bring them into something else,’ ” Reisner told me as she booted up her laptop to retrieve the records of the campaign. She offered to walk me through it, showing me just what they did.
First, they put up flyers at grocery stores and placed ads to recruit ordinary people for taste tests, and then they screened the applicants for people who already loved Dr Pepper. “These were our users,” Reisner said. In keeping with the company’s desire to deepen its base in communities that had growing African American and Hispanic populations, the resulting group of 415 tasters were spread among four cities: Los Angeles, Dallas, Chicago, and Philadelphia. Half were male. Six in ten were Caucasian. Their ages ranged between eighteen and forty-nine.
Cadbury wanted its new flavor to have cherry and vanilla on top of the basic Dr Pepper taste. Thus, there were three main components to play with. A sweet cherry flavor, a sweet vanilla flavor, and a sweet syrup known as “Dr Pepper Flavoring” that rendered the basic Dr Pepper taste. The precise ingredients in the final component remains a secret. In all, Dr Pepper is said to have twenty-seven ingredients. But besides water, the main ingredient, first and foremost, is sugar.
Moskowitz would pull together all his studies at Harvard, along with his math and the things he learned about taste and allure in his research for the Army and his many food company clients. He conducted what he calls an optimization, which translates like this: selecting the best element from a set of available alternatives. “What I say is, let’s base it on science,” he said. “Let’s make twenty or thirty or forty variations. When you do that, you’ll see that we like some of the variations more and like others less. And you can build a mathematical model that shows you exactly the relation between what’s under your control and how consumers respond. Bingo. You engineer the product.”
Engineering Cherry Vanilla Dr Pepper was no easy feat. Finding the bliss point required the preparation of sixty-one distinct formulas—thirty-one for the regular version and thirty for diet. (They differed ever so slightly in their ratios of flavorings.) The formulas were then presented to the tasters, who had to be managed a bit to get the most accurate results. Now and then, someone chooses to lie, usually just to rush through the tasting. But Moskowitz’s system is specifically designed to engage the tasters and convince them of the seriousness of the test. “We don’t let them talk,” Reisner said. “The rooms look professional, with nice computers. These are not junky places we have. The people are paid well, and the moderator will tell them there is no talking or discussing the products. They have to turn off their cell phones. They start to feel like their opinions count.”
Starting on July 12, 2004, in Los Angeles, Dallas, Chicago, and Philadelphia, the Dr Pepper tasters began working through their samples, resting five minutes between each sip to restore their taste buds. After each sample, they answered a set of questions: How much did they like it overall? (0=hate; 100=love.) How strong is the taste? How do they feel about the taste? How would they describe the quality of this product? And perhaps most important of all: How likely would they be to purchase this product? (From “definitely would buy” to “definitely would not buy.”) The scores were then added up. A score of 60 suggests the product will sell well. Fourteen of Moskowitz’s variations scored 61 or better, with two at 67 and two at a spectacular 70. Moreover, more than half of the panelists said that they would definitely buy the product, which in food marketing surveys is considered a terrific result.
The data that Moskowitz compiled in assessing the variations would outlive the cherry vanilla flavor itself. By evaluating consumer tastes so thoroughly, he had created a framework the soda maker could use to turn out a string of additional flavors that targeted specific groups of consumers. His data—compiled in a 135-page report for the soda maker, chock full of detailed charts and graphs—shows how people feel about a strong vanilla taste versus weak, various aspects of aroma, and the powerful sensory force that food scientists call “mouthfeel.” This is the way a product interacts with the mouth, as defined more specifically by a host of related sensations, from dryness to gumminess to moisture release. These are terms more familiar to people in tasting wines, but the mouthfeel of soda and many other food items, especially those high in fat, is second only to the bliss point in its ability to predict how much craving a product will induce.
In addition to taste, the consumers were also tested on their response to color, which proved to be highly sensitive. Reisner clicked on page 92 of the report, which showed a bright blue line streaking across a graph that depicted their liking for the color. “When we increased the level of the Dr Pepper Flavoring, it gets darker and liking goes off,” she said. The data can also cross-reference these preferences by age, gender, and race. The biggest surprise for most of Moskowitz’s customers relates to the consumer’s bliss point for sugar. The term bliss point, Moskowitz discovered, is actually a misnomer. It’s not a single point at all. It’s a range of points, which can be conceptualized like this: Take the graph that shows a bell curve, or an upside-down U; the top is actually a plateau with a range of points that will generate the same dose of pleasure. For Dr Pepper, the significance of this discovery is financial. Through the taste tests and mathematical modeling called optimization, Moskowitz found that Cadbury didn’t have to use the version of its new soda with the most syrupy Dr Pepper Flavoring. The same level of bliss could be achieved by using a bit less flavoring in each twelve-ounce can.
Page 83 of the Moskowitz report puts this phenomenon into a handy graph, and I followed along in my own copy of the report as Reisner explained. A thin blue line represents the amount of Dr Pepper Flavoring needed to generate the maximum appeal, and the line is not straight. It arcs, just like the bliss point curve that Moskowitz studied thirty years earlier in his Army lab. And at the top of the arc, there is not a single sweet spot but instead, a sweet range. The potential savings is merely a few percentage points, and it won’t mean much to individual consumers who are counting calories or grams of sugar. But for Dr Pepper, it adds up to colossal savings. The more soda it sells, the more money it can save by edging back on its key ingredient, the sugary Dr Pepper syrup, without losing the bliss.
“What we were able to do is show them that they could drop
the Dr Pepper Flavoring, and that saves them money,” Reisner told me. Instead of using 2 millimeters of the flavoring, for instance, they could use 1.69 millimeters and achieve the same effect. “That looks like nothing,” Reisner said. “But it’s a lot of money. A lot of money. Millions.”
In the end, Cadbury not only made its fall 2004 deadline for launching the new flavor that Moskowitz optimized. The launch was deemed a resounding success. “Don’t chug, people,” the company warned fans on its website. “You’ll want to savor this rich soda fountain goodness. So pull up a seat and enjoy a taste so flavorful it just goes on and on, and on.…”
By 2006, the company’s CEO, Todd Stitzer, was crowing to investors that the new flavor was not only a big hit among Peppers, it was also bringing newcomers to the brand who had begun to expand the soda’s reach from its heartland, the eleven southern states where more than half of the traditional Dr Pepper was being drunk by just 20 percent of the U.S. population. “We know that Dr Pepper’s franchise with consumers is rooted in its unique bold taste,” he told Wall Street analysts that year. “Cherry Vanilla Dr Pepper, launched in October 2004, builds on this heritage. The launch has surpassed ours and anybody’s expectations.” Cadbury was so enthused, he added, it would soon launch another new flavor called Berries and Creme, but he cautioned the analysts to be careful. Cadbury would be serving this flavor along with a newly released candy bar version of its classic Easter treat, the cream egg. For none of its products, it is safe to say, is the bliss point for sugar anything but maxed out.
