Disrupter – Stephen Friend

The driving force behind a genomics technology is conducting a second experiment: figuring out whether he can transplant the energy of a startup into the giant that bought his company.

Three times in his life, Stephen Friend has made what he calls a “great leap into the unknown.” Twenty-five years ago, he was a philosophy student in Indiana who decided to become a medical doctor. In the mid-1980s, after three years of pediatric training in Philadelphia, he moved to Boston and remade himself into a leading cancer researcher. And in 1996, he put away his lab coat to become president of Rosetta Inpharmatics Inc., a tiny Seattle-area genetics company with vast ambitions. In 1996, he put away his lab coat to become president of Rosetta Inpharmatics Inc., a tiny Seattle-area genetics company with vast ambitions.


Step into Friend’s spartan office at Rosetta today, and it takes a moment to realize how audacious he really is. He is a stocky man with a slightly untucked shirt and standard-issue chinos. Only when he starts to explain his career — and his goals that still lie ahead — do his energy and passions tumble forth.

“Any time you make a switch, there’s this zone of chaos,” Friend says. “You realize that everything you’ve done so far in your life adds up to maybe just one-twentieth of what you need to know going forward. All of a sudden, you’re about to start something completely different.” For other people, that’s terrifying. But not for Friend. “I’m happiest when I’m jumping blindly into a new area, going mostly on faith that it will work out somehow,” he says. “I find it tremendously exciting.”

So far, his gamble with Rosetta looks like a big winner. The Kirkland, Washington company is only five years old, but it has already become a landmark enterprise in the crusade to make sense of the human genome. Researchers at Harvard and Stanford and more than a dozen pharmaceutical and biotech companies now use the Rosetta Resolver Gene Expression Data Analysis System to analyze torrents of genetic data — fast. And anyone who wants to swap gene-expression data online will likely use Gene Expression Markup Language, a free, public standard developed in large part by Rosetta.


Earlier this year, in a world in which so many bold young companies’ hopes had been lofted — and then crushed — by the stock-market roller coaster, Friend and his fellow directors at Rosetta secured a safe, permanent home for their work. They negotiated the sale of Rosetta to one of the pharmaceutical industry’s most dominant players — Merck — for $620 million. Two big parts of the deal: keeping Rosetta as a stand-alone business in Washington State, and making Friend a vice president for basic research at Merck.

The Making of a Maverick

It would be easy for Stephen Friend to coast a bit, enjoying the $18 million from his Rosetta stock holdings. But so far he hasn’t even upgraded the Formica-topped furniture in his office. Instead, he is fascinated by the newest zone of chaos: a chance to help an organization as huge as Merck shake things up.

“What we’re discovering about the human genome has implications in every stage from preclinical research to the point where you start putting new drugs into people,” he says. “It’s hard to imagine a company with greater resources and possibilities than Merck. I don’t know how this will work out, but that’s part of the excitement.”


That’s classic Stephen Friend, say people who have known him for many years. “He has always been on the cutting edge, eager and impatient,” says Dr. Stephen Sallan, chief of staff at the Dana-Farber Cancer Institute in Boston. The two men worked closely together in the mid-1980s, when Friend was finishing a fellowship in hematology and oncology at Children’s Hospital Boston and Sallan was supervising Friend’s work.

Ironically, Friend grew up in a family that cherished the arts rather than the latest scientific breakthrough. His mother taught music theory at the Juilliard School in New York; his father was an acoustic engineer who also worked at Juilliard. Friend became a philosophy major at Indiana University and thought that he might make a career in medical ethics. But he decided that the only way he could truly understand the life-and-death issues that doctors face every day was to complete full medical training.

Then, partway through pediatric residency training at the Children’s Hospital of Philadelphia, there came a turning point. Working round the clock with cancer patients, Friend met dozens of children who were stricken with awful, inexplicable diseases. Their anxious parents would always ask the same two questions: “Why does my child have this disease?” and “What can you do to treat it?” All too often, doctors — Friend included — couldn’t give any useful answers. “You get such a feeling of emptiness,” Friend recalls. “Spend time in pediatric oncology with these children and their parents, and what you see will drive you for the rest of your life.”


Rather than trying to fight cancer one patient at a time with painfully inadequate tools, Friend moved to Boston in 1984 to pursue cancer research in hopes of one day benefiting many thousands of patients. In 1986, he scored a spectacular success when he isolated the gene associated with retinoblastoma, a fierce form of eye cancer. Working at Harvard Medical School, Massachusetts General Hospital, and Children’s Hospital Boston, he entered Boston’s research elite.

Yet once again, Friend felt frustrated that he wasn’t making more headway. “Knowing what was going on wasn’t the same as being able to treat it,” he recalls. What’s more, Friend began to worry that Boston’s enormous concentration of medical talent might be stifling advances instead of furthering them. “Everyone in Boston feels certain that they are working on the right next step in research,” he later explained. “So if you say, ‘I think I see a better way,’ it’s very hard. The natural impulse of everyone else is to say, ‘I had that thought a few years ago, and it doesn’t work.’ “

All the while, job overtures were coming from Seattle. Lee Hartwell, a world-renowned geneticist and yeast researcher, was building up a genetics team at the University of Washington. Friend admired Hartwell’s work and began to regard the nearby Fred Hutchinson Cancer Research Center as a place where mavericks could thrive. Before long, Friend was in Seattle for a one-year sabbatical at the Hutch. He then stayed on as director of molecular pharmacology there and also became professor of pathology at the University of Washington.


From Scientist to CEO

Friend’s timing was perfect. The new field of bio-informatics was coming of age. Scientists were using sophisticated computer software to analyze a growing mountain of genetic data about yeast, mice, and, eventually, people. The Seattle area, home of Microsoft, was a hotbed of innovation. And a second breakthrough was occurring in the research lab of Leroy Hood, a genetics researcher who had just moved to the University of Washington.

One of Hood’s scientists, Alan Blanchard, had devised a way to synthesize as many as 25,000 snippets of DNA onto a single laboratory slide, where they could be fixed neatly in place and used to analyze patterns of gene expression. Scientists had long wanted such a powerful analytical tool, but they had no idea how to build it without taking years or millions of dollars. It was Blanchard’s genius — or madness — that inspired him to build such a tool by taking a standard Epson ink-jet printer, dismantling the printing head, loading it up with the building blocks of DNA, and then squirting DNA bases onto the slide with astonishing precision.

“All of a sudden, we had the potential to analyze everything that was going on inside a cell,” Friend recalls. “This wasn’t going to be just a cool little idea for three or four guys in a lab. It would take hundreds of people to do it. Academia couldn’t be the right setting for it.” Friend, Hartwell, and Hood began talking about the best way to combine various breakthroughs and exploit the combined opportunities. In late 1996, each of them realized: We’d better start a company.


Neither Hartwell nor Hood wanted to run the company full-time, so Friend stepped in as president. His parents — who had been so proud of his careers in medicine and in research — were dismayed. “They had always viewed business with suspicion,” he says. But Friend threw himself into this new calling with gusto. He recruited scientists and engineers. He wooed venture capitalists. And in late 1997, he lined up $17 million from backers led by OVP Venture Partners and Vulcan Northwest Inc., the venture vehicle of Microsoft cofounder Paul Allen.

For its first 18 months, Rosetta made hardly a ripple. Its proprietary software had to be written and debugged. Its gene-array technology had to be perfected — which was no small feat in Seattle, where high humidity periodically threatened systems that required virtually bone-dry conditions. From Hood’s lab, Blanchard joined the team and began ordering Epson printer heads by the dozen.

“Do you want any ink cartridges with those?” a salesman asked at one point.


“No,” the Rosetta scientists replied. “That won’t be necessary.”

Before long, Rosetta faced a crossroads. It could aim to be a high-volume producer of gene arrays itself and spend huge amounts of money setting up factories. Or it could concentrate mostly on cutting-edge research, licensing its technology to a bigger, more established company that would run the production lines. Friend decided that the latter choice was wiser. So he and his colleagues negotiated an alliance with Agilent Technologies, which now cranks out many tens of thousands of gene-array chips using Rosetta’s know-how.

In 1999, with the Agilent licensing deal generating the startup’s first few million dollars of revenue, Rosetta could pursue an even more exciting market: perfection of software that would help giant drug companies analyze a flood of data from the gene arrays. “We always felt that analytical software would be the best part of the value chain,” recalls Chad Waite, an OVP partner and Rosetta director. But Rosetta’s ambitions would go nowhere unless it could sign up pharmaceutical customers who would share what ordinarily would be top-secret data.


That’s where Friend opened doors faster. “Steve had the stature to talk to top pharma researchers as a peer,” Waite recalls. “He was incredibly enthusiastic about what Rosetta’s technology could do for them.” By last year, more than a dozen top-tier life-sciences companies and universities were Rosetta licensees. In all of those negotiations, insiders recall, Friend was the “good guy,” focusing on shared opportunities. Rosetta’s senior VP and COO, John King, played the role of the “tough guy,” making sure that Rosetta got the best terms possible from each deal. In early 2000, Rosetta made a successful public stock offering — and to almost everyone’s surprise, Friend handled the role of CEO as though he had been training for it all his life.

Big Ambitions, Big Backers

As Friend pushed for closer collaboration with big drug companies, Rosetta agreed to meet with executives from Merck. But when a nine-person Merck delegation arrived in Kirkland, Rosetta managers began whispering to each other: That team is way too big to send to talk about a research partnership.

Sure enough, partway through those talks, Merck executive Peter Kim took Friend aside and asked him, “What would you think if we talked about acquiring Rosetta?” Friend was suitably coy at first. But inside the Rosetta boardroom, he argued hard for the deal. From the moment he helped found Rosetta, he argued, his ultimate goal was to hasten the development of life-saving medicines. As he saw things, the Merck deal would give Rosetta unparalleled access to compounds, libraries, and facilities — enabling the little company to maximize its impact.


Meanwhile, Merck agreed to pay a 70% premium over Rosetta’s predeal share price. It promised to keep the Rosetta team in Kirkland, as a separate subsidiary that could still seek non-Merck clients for its software products and services. Friend would remain president of the Rosetta business — and take on powerful responsibilities at Merck as well.

Already, Friend’s new position has encouraged him to think even bigger. “For much of Rosetta’s existence, we’ve avoided cancer-related research,” he says. There are many factors at work in the course of any cancer, and the interactions can be maddeningly complex. But for the past few years, Friend says, two researchers in Amsterdam have been urging him to apply Rosetta’s tools to cancer anyway.

“They persuaded us,” he says. “So we are now working on developing genetic markers that could help us determine the prognosis for someone with cancer — whether or not they would be responsive to therapy. It’s exciting to see what could be an entirely new diagnostic tool.”


And now — 16 years after his early, disheartening encounters with cancer during his pediatric residency training — Friend can again dream of lifesaving breakthroughs. “This will play out over many years,” he warns. “But if we’re lucky, there may be new ways to start identifying the targets that really matter for fighting the disease.”

George Anders ( is a Fast Company senior editor based in Silicon Valley. Contact Stephen Friend by email (

Sidebar: What’s Fast

Over the past two decades, Stephen Friend has published more than 60 scientific articles and commentaries on topics ranging from breast-cancer screening to DNA analysis. But some of his best new insights apply to business leadership — and the special challenge of guiding a fast-growing startup to stay on track in turbulent times.


Dare to enter a “zone of chaos.” Exciting young fields like genomics seldom have clear rules. There’s no way for the pioneers to be experts when they start. But the people who jump in first are likely to find the best opportunities.

Don’t get distracted. Startups can be seduced by a multitude of opportunities — and companies that try to do too much may end up accomplishing nothing. When Rosetta developed its own gene-array technology, it could have decided to build such arrays by the hundreds of thousands. But it didn’t, because Friend wisely believed that the young company needed to pour all of its energy into software development and research breakthroughs.

Cultivate a few great customers. In its first three years, Rosetta didn’t hire a single sales representative. Instead, top executives led by Friend hit the road themselves and courted research chiefs at the likes of Amgen, DuPont, Merck, and Monsanto. After winning just a few of these accounts, Rosetta was in a much better position than it would have been if it had decided to chase dozens of smaller contracts with less-renowned partners.