How To Innovate
How Mars Built a Business
"There is little reason for an individual to have a computer in their home," Ken Olsen, the president and founder of the Digital Equipment, famously said in 1977. As Olsen's quote suggests, predicting demand for new, innovative products and services can be difficult, in part because many of the traditional methods of market testing—using historical data to forecast sales, for instance, or asking customers in a focus group to compare a new product with an existing, competing one—aren't well-suited to the innovation process. This was the dilemma that Dan Michael, then R&D director for Mars' M&Ms brand, faced in 2000. He and his research team at the advanced R&D lab in Hackettstown, N.J., had an idea: to make customizable M&Ms printed with the word or image of a customer's choosing. So they began experimenting with new printing technologies. Initially hacking together prototypes using printers bought from Staples, they had settled by early 2003 on the best printing technologies for the task, applied for patents, and taken their concept to Mars senior management. "Everyone thought the idea of customized M&Ms was phenomenal," says Jim Cass, now general manager of Mars Direct. But it was just that—an idea. The R&D team didn't have a marketing strategy or even customer research showing that customized candies would be a viable business, so there was some skepticism. Michael and team needed to convince management there could be a market for customized candies. To do that, they had to reinvent the development process—and the role of marketing within it. Mars had recently launched an innovation initiative called Pioneer Week. Select research teams were given a modest budget and 90 days to build a trial production line, after which the new product would be made available to Mars' 65,000 employees. "The teams were allowed to bypass some of the testing normally associated with product development," says Marc Meyer, a professor at Northeastern University's College of Business in Boston, who has studied the company. "A Real Skunkworks"Michael's team, still fewer than a dozen people, was selected at Pioneer Week, and in June 2003—after three hectic months—their product, then called M&M's Colorworks Print Shop Milk Chocolate Candies, went on sale. They received orders for 800 pounds of custom-printed white M&Ms (the only color offered initially) within the first day. "It was still a real skunkworks operation—we had one small printer, and we hand-bagged everything," says Cass. "But the strategy was, 'make a little, learn a little; make some more, learn some more." The internal trial provided critical marketing feedback: The four-pound minimum order size was too big, and customers wanted colored candy and "party favor" packaging options. Setting the price was another challenge. For the internal launch, the team chose $12 a pound, $4 more than the retail price for standard M&Ms. Since then, according to Cass, the price has changed four or five times. "We were learning incredible amounts of information on a daily basis and continued to optimize the price/value equation," he says. Today consumers can buy one seven ounce bag of customized M&Ms for $16.99. Larger orders cost less per bag, with a 10-pound bulk order going for $250. Customers, who often order the customized candies for weddings or milestone events, have proved willing to pay the premium. Learning From the CustomersReady to take it to the next level, the team began selling My M&Ms, as the custom candies are now known, to the public through a small link from the main M&M Web site in March 2004. Without any marketing blitz, sales took off. That's also when the team began to do more serious customer research. "We reached out to every [My M&Ms] customer, and roughly 25% responded, providing detailed demographics as well as information on why they had made the purchase," says Cass. Because the customer feedback relates to actual purchases, it is more reliable than answers from traditional focus groups, where people talk hypothetically about why they would make a purchase. In 2006, Mars' My M&Ms experiment became a formal business unit called Mars Direct. Famously secretive, the company won't share sales data, although Meyer, the Northeastern professor, wrote in his 2007 book, The Fast Path to Corporate Growth, that soon after the public launch "sales had surpassed $10 million and continued to accelerate." The product was launched in Europe in 2007 and will soon be introduced in Australia. What can executives learn from Mars' approach to marketing an innovative product? Think Iteratively, Not Linearly Give up the idea that you need to have all the answers about product features, packaging, target market, and so on before a product launch. Instead, run a small trial of the actual product or service, learn from it, revise, repeat. Forget Focus Groups When it comes to new-to-the-world products or services, don't rely on what customers say they think or want. As Henry Ford is quoted as saying, if he'd asked his customers what they wanted, they would have said a faster horse. Be Your Own Test Market Big companies can gain invaluable information, at far less than the cost of hiring a market research firm, if they treat their employees as a test market. Smaller companies can look to friends and family.