King's critics are less kind: "We see King like remora fish stuck on the side of the shark," says Derek Lowe, author of industry blog In the Pipeline. "We see these people as living off the fruits of our labors."

Jolly takes such criticism in stride. Collaboration is the soul of R&D at King. In fact, Jolly says the company has had to attract a "peculiar type of animal" to the department. On one hand, "We didn't want people who would just throw things over the wall [to a CRO or other partner] and not take ownership for it," he says. But, it also was important to find "people who knew the contracting business and knew how to get maximum productivity out of the people you contract with—people who know how to recognize the signals as to when things are going awry." Not surprisingly, about half of Jolly's staff held previous positions in contract services businesses.

King R&D's non-possessive attitude is what's allowed the department to function "like a central processing unit in a computer," says Jolly. "We have every function that you'd find at a pharmaceutical company"—preclinical, discovery, pharmacology, toxicology, etc. "The difference is, we don't have the operational overhead." Some of King's scientists, through affiliations with noted experts in their fields, work collaboratively in academic labs. But mostly, scientists at King leave their white coats at home, sit at desks, and function like project managers.

Jolly doesn't see King's take on R&D as scientifically compromising. "We still control the science," he says. "We make sure that we design the plan and the scope. We rely on others to execute."

And financially speaking, King believes its approach only makes sense. The company recently underwent a benchmarking process that examined its R&D department with respect to time and cost. It uncovered comparable timelines with the rest of industry, but substantially lower costs.

"That's why we are running 12 or 13 programs with 62 people," says Jolly. "In our development, our key mantra is to do only the minimum number of studies necessary to get a drug approved, to plot a critical path, and to outsource as much of it as you can, as quickly as you can." Jolly boasts about one of his clinical scientists, who currently is in charge of four major programs. "If you were to go into industry, you probably would never find one person running four major programs."

Perhaps not today, but the economic pressures of the next few years seem likely to drive other companies to consider King's achievements in virtual R&D, even if few of them attempt to make it the centerpiece of their product-development strategy. There's little doubt that King is making strides with something that will soon be seen as a key pharma skill set. But will it be enough to create a sustainable business with growth potential? Can a virtual approach rise to the new data demands posed by payers and others? Will there be enough late-stage products on the market to feed King's pipeline? The "remora" companies like King have had an interesting place in pharma. Increasingly, though, they're going to face the challenge King is grappling with right now: Get a new grip and hang on for the ride—or learn to swim alone.