Lilly’s RD Prescription

How does pharmaceutical giant Eli Lilly accelerate the pace while reducing the cost of innovation? By corralling scientists around the world in a Web-based system of eRD.


Pharmaceutical companies often face a daunting R&D challenge. When a blockbuster drug loses its patent protection, the drain on revenue and profit is immediate. Meanwhile, it takes a huge amount of time (12 to 15 years) and an awfully big budget (approximately $800 million) to bring a major drug to market. Finally, scientists don’t have many places to go for help. Most drug research is done in secret; researchers rely on the small, personal networks within their companies.


Eli Lilly and Co. was facing all of those challenges as it prepared to confront “Year X” (2001), the year that its patent for Prozac would expire. Executives knew that Year X would spark a major loss of revenue; at its peak, Prozac accounted for 34% of annual sales. The natural response, which Lilly pursued, was to hunker down and devote armies of scientists and vast amounts of money to the search for the next blockbuster drug.

But Lilly launched another experiment as well. It helped form a wholly owned subsidiary called InnoCentive LLC, which does what the name implies: It creates incentives for innovation. The InnoCentive Web site is designed to tap a global community of scientists who work on critical problems that chemistry-driven companies want to crack but can’t. InnoCentive awards as much as $100,000 to scientists who unravel a problem. In return, companies get the intellectual-property rights to the solution, which is often an “intermediate” of a new drug (a substance formed in the middle stage of a series of chemical reactions).

“We’re punching a hole in the side of the laboratory and exposing mission-critical problems to the outside world,” explains Darren J. Carroll, president and CEO of InnoCentive and an Internet veteran. “This isn’t sock puppets or Super Bowl advertising. It’s using the Net to communicate, collaborate, and innovate.”

Lilly’s development team dubbed early iterations of the InnoCentive prototype “BountyChem,” and its software logo initially depicted chemical compounds surrounded by a bull’s-eye. Although the team was building a Web site for the 21st century, its members were inspired by a 19th-century business model — that of the bounty hunters of the Wild West.

Here’s how the model works: Drug companies, called “seekers,” go to the site and put up a kind of “Wanted” poster that describes the target they’re trying to nail. Bounty-hunter scientists, called “solvers,” sign a confidentiality agreement and then go to a secure “project room,” which contains data and product specifications related to the problem. Like the bounty hunters of old, solvers take on much of the risk in hopes that they will be able to earn a reward.


By harnessing the Web, InnoCentive has built a virtual community of scientists that puts the organization “within one or two degrees of separation from every single organic chemist in the world,” says Carroll. So far, 7,000 scientists have registered at InnoCentive, and there are 2,400 project rooms in use, organized around 33 problems.

As of early 2002, Werner Mueller, now an independent chemist and formerly head of technology for the specialty-chemicals group of pharmaceutical giant Hoechst/Celanese (before it became part of Clariant), has bagged InnoCentive’s biggest bounty. He devised a two-step synthesis for the compound 4-(4-hydroxy-phenyl) butanoic acid, an intermediate for a drug that Lilly is developing.

Mueller earned a $25,000 reward, but he is not entirely happy. He argues that under the InnoCentive model, the solver assumes too much risk. “You have to be willing to work for three months and possibly come up with nothing,” he says. “The seeker should do more to share the risk. Right now, there’s high risk for the solver and zero risk for the seeker.”

Carroll concedes that solvers take on considerable risk, but that’s the point: InnoCentive cuts the cost of drug discovery by spreading the risk to outside scientists. “So now it’s up to solvers to decide whether the risk is worth the reward,” he says. “But ideally, they already know the answers. They’ve just been waiting for us to come along and ask the right questions.”

Carroll emphasizes that InnoCentive is still a work in progress. To succeed, it must attract more companies to its site. It also needs to increase the number of problems that are posted to the site, so enough solutions are generated to make the company self-sustaining. InnoCentive, which acts as a broker, derives much of its revenue from commissions paid by seekers whose problems are solved.


Still, Carroll is undeterred. He anticipates that InnoCentive will sign up a second seeker within weeks. And it is already adding new elements to its site, including a list of problems relating to drug formulations, analytical chemistry, and radiolabeling (a process in which a radioactive tag is added to a compound to track how it works in the human body).

Carroll believes that InnoCentive can win over much of the drug industry, reasoning that access to the best minds in the world is critical to the industry’s success. And he maintains that despite the minefield of concerns over risk and intellectual-property rights, InnoCentive will attract scientists — simply because it gives them a compelling glimpse of a different kind of future.

“When scientists look at us, they see a choice,” explains Carroll. “They can continue to work for large, R&D-based organizations, or they can become free agents and work for themselves. Under certain circumstances, they might even do both. Free agency has never been an option in the hard sciences — until now.”

Learn more about InnoCentive online ( For a collection of articles on innovation and the pharmaceutical business, click here.

Sidebar: Lab Partner

How does a pharmaceutical giant in Indianapolis tap into a global community of independent professionals? It turns to a small company in Indianapolis that taps into a global community of independent professionals.


Quovix LLC is a 10-person outfit backed by a virtual workforce of 455 free-agent programmers. It designed the software that enables InnoCentive to set up Web-based workspaces. How do you harness a workforce of 455 coders, most of whom you’ve never met? “Simple,” says Marty Morrow, CEO of the two-year-old startup. “I focus on the three phases of virtual-product development: water, slush, and ice.”

The water stage occurs when Quovix lands an assignment. Morrow’s team posts the project to Quovix’s Web site, where developers click in with feedback. “Hundreds of people provide input and ideas,” says Morrow.

If Quovix wins the contract, it enters the slush stage. “We define the project down to a gnat’s eyelash and put it out for more feedback. Then we put a price tag on the project, and we run a reverse auction. The person with the most experience and lowest bid wins.”

Once the project is awarded, it enters the ice stage. This is when the work gets done. “We have tighter control over the developers in our community than most companies have over staff developers who are sitting 20 feet away,” says Morrow. “We run daily test patterns on their work, and we know within a day if we’ve got a problem.”