September 13, 2004
(Computerworld)
Aided by the Internet and advanced search techniques, a handful of companies is leading a revolution in the way new products are designed and developed. By looking beyond their own R&D labs -- to suppliers, universities, freelance inventors and even competitors -- they are accelerating the pace of innovation while sidestepping the costs and risks of in-house research.
"The R&D model that most companies are following is broken," says Larry Huston, vice president for research and development at Procter & Gamble Co. in Cincinnati. "There's a drive to increase innovation budgets beyond the [revenue] growth of the firm. That's not a sustainable business model."
But P&G, which spent $1.7 billion on R&D last year, has found a new model, called open-market innovation. Indeed, the consumer products maker has embraced the idea so enthusiastically that it no longer refers to its product-innovation process as R&D; it's now C&D, for "connect and develop."
P&G makes the connections among its employees and external sources with a variety of tools and techniques, including an intranet, Web sites, commercial and homegrown search engines as well as several "innovation networks" -- intermediary companies that match innovation seekers and suppliers. In just two years, the company has boosted the percentage of product innovations that come from outside sources from less than 20% to 35%. P&G's CEO wants to raise that to 50%.
"This is a classic application of the Internet, going back to its origins," says Darren Carroll, CEO of InnoCentive Inc., one of the innovation matchmakers. "For those of us on Arpanet in the beginning, it was all about scientists and engineers sharing problems and solutions."
Innovation Networks
The following are some of the resources P&G uses to connect and develop:
• InnovationNet. InnovationNet is an intranet Web portal for 18,000 P&G innovators in R&D, engineering, market research, purchasing and patents. Nabil Sakkab, a senior vice president for R&D, calls it a "global lunchroom" for the exchange of ideas.
• InnoCentive. Founded by Eli Lilly and Co. but operating independently, Andover, Mass.-based InnoCentive claims to be the "largest virtual laboratory in the world." It posts scientific problems from its 30 "seeker" members to a proprietary network of 70,000 registered "solvers" around the world. Each posting includes a promised cash award for the solution. (The problems, which are well defined and typically have provably correct answers, can also be viewed at www.innocentive.com/servlets/project/ProjectInfo.po). "The success rate so far has been around 50%," Sakkab says. "Not bad for problems we failed to crack in-house."
• NineSigma Inc. This Cleveland-based firm helps its 50 or so clients prepare technical briefs describing projects or problems they are trying to solve and then sends the briefs -- without identifying the originating companies -- to thousands of researchers around the world. The idea is not to get back specific solutions, as InnoCentive does, but to identify people most likely to be able to provide solutions on a contract basis.
• YourEncore Inc. An Indianapolis-based network of about 400 retired scientists and engineers, YourEncore was created 10 months ago by P&G and Lilly but now operates independently. It matches its members with clients for specific, short-term job assignments and pays them their salaries at retirement plus 20% .
Risk-Free R&D
Indianapolis-based Lilly, another pioneer in open-market innovation, supplements its internal R&D resources with the services of InnoCentive and YourEncore. Alph Bingham is a vice president at Lilly Research Laboratories and InnoCentive's chairman. He says that a company like Lilly traditionally would look on the outside for the very best person -- perhaps even a Nobel laureate -- to solve a hard research problem. But it might pay that person $50,000 for six months of work and still not get a solution.
"Betting on the Nobel laureate is the best way to go if I have to bet on one person," Bingham says. "But what if an assistant professor at Morgan State University is actually better qualified to solve that particular problem, at that moment, than anyone else on the planet? But I can't find him, so I use a mechanism like InnoCentive."
InnoCentive pays its award money -- typically between $5,000 and $100,000 per problem - only when someone comes up with a workable solution. "So now the second reason for going this way manifests itself," Bingham says. "I don't pay the $50,000 until I know the problem was solved."
The third advantage of going with an external innovation procurer is that doing so preserves the anonymity of both buyer and seller, so there's an intellectual property firewall between both parties. "It gives a company like Lilly the ability to hide its problems in plain sight," Bingham says. "Someone can look at our problem and not know it's a drug application. It could be a pesticide application."
Search Magic
Cutting-edge search technologies are essential to the connect-and-develop approach. NineSigma creates a unique database of potential respondents for every client request. "The databases are generated through a variety of searching techniques, some of which are proprietary," says Shauna Brummet, vice president for operations at NineSigma. "It goes significantly beyond Google, and the techniques are evolving."
For each problem, NineSigma sends out 6,000 bid requests on average and receives 10 to 100 responses. Getting high-quality responses is a key to success, says Richard Swarz, chairman of NineSigma, and that requires sophisticated search algorithms to build just the right mailing list, as well as carefully crafted requests for proposals.
InnoCentive searches its database of 70,000 solvers and sometimes other sources as well, depending on the problem. "We have developed some proprietary algorithms that can identify with a higher probability the folks most likely to participate in a challenge," says InnoCentive's Carroll.
P&G is seeking to identify what it calls "superconnected giants" in the networks. "They are connected through patent literature, they are publishing a lot, they speak at conferences, maybe they are at the center of a hub as department head at a major research hospital," Huston says. The average researcher knows 2,000 people; the superconnected ones that P&G covets know 10,000 people, he says.
The Final Connection
While companies like P&G are beginning to tap into open-market innovation, the various methods of doing so are not yet well integrated, says Navi Radjou, an analyst at Forrester Research Inc. "The laboratory management systems and discovery tools that scientists use must provide seamless integration with things like InnoCentive," he says. "I think you'll see that happen in the next two or three years."
But Lilly isn't waiting for software vendors to step up to the challenge. It recently launched a project to automate the internal processes surrounding the use of InnoCentive and is also developing interfaces to InnoCentive's own workflow. "We are doing the IT design for an internal portal to these [third-party] systems so that it becomes part of the scientist's natural workflow," Bingham says.
For example, a Lilly scientist today seeking help with a problem would first have to know InnoCentive exists and then find someone at Lilly to explain how it works, fill out approval forms and disclosure forms, contact the relevant person at InnoCentive and so on, says Bingham. With new workflow automation and interfaces to InnoCentive, that scientist will be able to accomplish the same things with a few mouse clicks. And once the problem definition has been posted by InnoCentive, the Lilly scientist will be able to track its progress online.
Carroll says open-market innovation will dramatically expand in scope over the next five years. "You will see it expand into statistical analysis of business problems, graphics design, advertising and other services industries where this model may apply even more strongly," he says.
Asked about the future of open-market innovation at P&G, Huston says, "The current business model that people are following is not sustainable, and more and more companies will move this way. It's going to become even more important as we face a scientific talent shortage in the U.S.
"This is the future," he says. "People just don't realize it yet."