Part IThe Story
Ten Minutes on a Friday
He asked for ten minutes. That much everyone agrees on.
It was early 1976, January by most accounts. The scientist was Herbert Boyer, a biochemist at the University of California, San Francisco, whose work with Stanley Cohen of Stanford three years earlier had shown that DNA could be cut from one organism and spliced into another. Boyer was not looking for a business partner. He did not, by his own admission, know what the young man on the phone did for a living. "I didn't know what venture capitalist was in those days," Boyer recalled decades later. What he understood was the part about money: "laboratories always needed money, still do." So he said fine, come by. Late Friday afternoon. Around five o'clock.
The young man who walked through the door was twenty-eight years old, or twenty-nine. The record wobbles on this point, as it wobbles on several others in the life of Robert A. Swanson, as though the man moved too quickly for the archive to keep pace. He was also unemployed. That last fact is the one the commemorative accounts tend to leave out, and it may be the most important fact in the room.
— Herbert Boyer, recalling his first meeting with SwansonHe came to my lab around five o'clock, walked in with a suit and tie, you know, and other than the suit and tie he looked like one of my students, he was very young.
The suit and tie is what Boyer remembered. In a university lab at the end of the week, among graduate students in whatever graduate students wore in San Francisco in 1976, a man in a suit is either a lawyer, a salesman, or a person who has dressed for a job interview because he has been going to job interviews almost every day. Swanson was that third thing. He had been let go by the venture firm Kleiner & Perkins at the end of 1975, and he was interviewing nearly daily for work while, in the margins, cold-calling the scientists whose names he had found in the literature on recombinant DNA. Most of the calls, by most accounts, went nowhere. Boyer's went somewhere.
Ten minutes became three hours. At some point the conversation left the lab for a bar, and at some point beers were involved, and by the end of it the two men had agreed to put up $500 apiece to cover the legal fees on a partnership whose purpose was to turn the most controversial technique in biology into medicine for sale. Here the legend splits again. The citation accompanying Swanson's National Medal of Technology places the whole ten-minute meeting "at a San Francisco bar." Boyer places it in his lab. Genentech's own corporate history says only that Boyer "agreed to give the young entrepreneur 10 minutes of his time." I cannot tell you which is right, and I am not sure it matters; origin stories of this magnitude acquire a geography of their own. What matters is the clock. A man with nothing — no job, no lab, no product, no company — asked a stranger for ten minutes and got three hours, and the difference between those two numbers is roughly the size of the biotechnology industry.
"Here comes this brash young entrepreneur filled with enthusiasm and ideas and ready to go," Boyer said later. "I recognized right away that he had the drive and the understanding."
Within a year, the pair would make a human hormone inside a bacterium. Within three, human insulin. Within five, they would stage what was then the largest initial public offering in American history. And Swanson, who had walked into that lab with a suit and a request, would spend the next two decades running the company he had talked into existence — impatient with everyone, patient with the one thing that counted, and dead at fifty-two of a cancer his own company's medicines would one day be approved to treat.
By the Numbers
From a $1,000 Partnership to an Industry
$500Each founder's stake in the original partnership, to cover legal fees
$100,000Kleiner & Perkins' first investment, from an $8 million fund
10,000 sq ftGenentech's first lab: a leased section of an airfreight warehouse, February 1978
$35MRaised in the October 14, 1980 IPO — then the largest ever
$2.1BPaid by Roche for 60% of Genentech in 1990
$46.8BPaid by Roche in 2009 for the shares it did not already own
52Swanson's age at his death from glioblastoma, December 6, 1999
The Son of the Shift
Robert Swanson was born on November 29, 1947, in Brooklyn, to Arthur J. Swanson and Arline Baker Swanson, and he did not stay in Brooklyn long. While he was still an infant the family moved to Florida, where his father went to work for Eastern Air Lines as the leader of an airplane electrical maintenance crew. Arthur Swanson worked in shifts. That is nearly all the public record says about him, and it says a good deal: a man responsible for the wiring of aircraft, a man whose hours were set by the schedule of other people's flights, a man whose job, at its most essential, was to make sure the planes stayed in the air.
The household creed was generational ascent. Swanson said he was taught from an early age that his generation would do better than the one before it, and the instrument of that betterment was to be a college degree — he would be the first in the family to earn one. The family had a particular institution in mind. When the Massachusetts Institute of Technology admitted him in 1965, the pride at home was, by his own telling, considerable.
He majored in chemistry. It did not take. The moment of clarity came not in a lecture hall but in a summer job, the kind of entry-level industrial posting that is meant to confirm a vocation and sometimes instead dissolves one.
"At the end of my junior year, I... got a summer job working for a chemical company," he said in a 1996 interview. "One of the things I discovered was that I enjoyed people more than things. So I said, 'Gee, this probably isn't going to be what I'd want to do all my life.'"
Gee. It is worth sitting with the plainness of that, the Florida-boy register of a decision that would, in the fullness of time, put a chemist's training to work in a world that did not yet exist. He did not drop chemistry. He did something cleverer and more characteristic: he petitioned MIT to let him take the first year of courses at the Alfred P. Sloan School of Management while still an undergraduate, and MIT agreed. In those graduate courses he found the two subjects that would occupy him for the rest of his life — organizational development, and the commercialization of innovative ideas. He left MIT in 1970 with a bachelor's degree in chemistry and a master's in management, both conferred, depending on whose records you trust, in 1970 or a year apart. He was a member of Sigma Chi. He was, by every later account, gregarious.
There is a pattern here that would repeat. Swanson did not reject the path laid out for him so much as negotiate a detour that kept the destination intact. The family wanted MIT; he got MIT. The degree was supposed to be chemistry; he got chemistry, plus the thing he actually wanted. Years later, at Genentech, he would make the same kind of trade with academic scientists — you can keep your publications, he would tell them, after the patent is filed — and with his venture backers, and with Eli Lilly, and eventually with Roche. Everyone got something they wanted. Swanson got the thing that mattered.
Near the end of his life, Charles Vest, then president of MIT, recalled that Swanson would ask him about the students the Institute was attracting, and would always come around to the same question: "What's going to happen to the average kids like me?" Vest called it humility. It may also have been accuracy, or at least memory — the memory of a boy whose father worked shifts, who learned in a summer job that he wasn't going to be the best chemist in the lab, and who decided he would rather be the person who got the best chemists into the lab in the first place.
A String of Failures
The usual biography of a founder bends toward inevitability. Swanson's resists it. Before Genentech he was, by the frank account of the man who hired him, a venture capitalist who did not do very well at venture capital.
He started at Citibank in New York, managing a venture investment group, and did well enough that he and a colleague were chosen to open a San Francisco office for Citicorp Venture Capital. He moved west in 1973 — which put him, as two of his earliest scientific recruits later observed, "only minutes from his office" from the labs where Boyer, Cohen, and their colleagues were splicing genes. Geography is not destiny, but it helps.
The San Francisco investments, however, did not do well. One in particular, a science-based company called Antex, went bankrupt. Swanson was assigned to claw back what money he could from the wreckage, and the person he worked alongside in that salvage operation was Eugene Kleiner, cofounder of the partnership Kleiner & Perkins.
Kleiner had already lived through the founding of one industry. He had been a cofounder of the venture firm whose name, along with his partner's, would become a kind of Sand Hill Road liturgy, and he was a man known for a particular kind of judgment. What impressed him about Swanson, Tom Perkins later said, was "Swanson's ability to think straight and get things done." A bankruptcy, in other words, served as the young man's audition. Swanson himself later described Antex as a lucky break — the failure that introduced him to the people who would back his success.
He joined Kleiner & Perkins as an associate in late 1974 — the firm then occupied the top floor of Embarcadero Center Two — and things went poorly again.
"Bob worked on various deals for us," Perkins recalled. "And, I have to say, not very successfully. Eugene, in particular, was increasingly unhappy with Bob. He didn't think much was going to be happening."
What Swanson spent his time on, rather than closing deals, was recombinant DNA. His interest had been sparked at a lunch with the Nobel laureate Donald Glaser, and he directed the full force of that interest at one of the firm's portfolio companies, Cetus, trying to persuade its leadership to pursue genetic recombination projects. Cetus had been founded in 1971 — five years before Genentech, a fact that complicates every claim that Genentech was "the first biotechnology company" — but its focus was not gene-splicing, and its leaders judged the venture too risky. Kleiner & Perkins and Cetus parted ways. The falling-out, according to Sally Smith Hughes's oral history of Swanson, was one of the main reasons the partners advised their associate to look elsewhere. By the end of 1975, his position was terminated.
There is a strange symmetry in this that the commemorative accounts rarely dwell on. Swanson was effectively fired for being too interested in the thing that would make him, and his firing firm, famous. The partners did not want to be in the recombinant DNA business. A few months later they would write the first check to it.
Perkins, in the documentary Something Ventured, described the ambiguity of the arrangement: there was not much for Swanson to do at the firm, so Swanson said he would look into biotechnology, and Perkins told him that if something was there, they'd be interested — otherwise. And then, the recollection goes, Perkins trailed off into a gesture.
Otherwise. The young man in the suit, cold-calling scientists from a list, was living inside that unfinished sentence.
Bob Swanson and a Checkbook
The partnership was worth $1,000. It had no employees, no lab, no product, and a name — Genentech, compressed from Genetic Engineering Technology — that its founders would later insist, with more confidence than accuracy, described the world's first company of its kind.
Swanson then made the decision that the rest of his life would hang on. He had been interviewing for jobs. He stopped.
— Robert Swanson, on deciding to pursue Genentech full-time rather than take a jobLook, I think this is important. If I don't do this, I'm not going to like myself so much for not having given it a shot.
It is a modest sentence for an immodest act. Not I knew it would work. Not I saw the future. Just the calculus of regret, worked out by a twenty-eight-year-old who had already failed twice and had decided that a third failure on his own terms was preferable to another success on someone else's.
He rented a small office in downtown San Francisco and began, with Boyer, to write a plan. Boyer drafted the science. Swanson drafted the market. The question was what to make first, and Swanson's answer arrived quickly: human insulin. The logic was a braid of three strands. Scientifically, insulin was well characterized, its structure already worked out, which made it easier to attack in theory. Medically, the insulin then in wide use came from pigs, and many patients developed allergic reactions to it; a human version would presumably avoid that problem. Commercially, world insulin sales already exceeded $100 million a year and were growing. Boyer agreed. By March 1976 Swanson had a business proposal in hand, and he took it, of all places, to Kleiner & Perkins.
Tom Perkins was a man who had a theory of risk and would tell you about it. "Everybody knows that venture capital should never openly fund pure research," he said in Something Ventured. "My idea in everything has been to put the risk up front and get rid of the risk as fast as you possibly can." What Swanson and Boyer brought him was, in his assessment, pure research — a plan to raise roughly $3 million, build a facility, hire the people, and see if it would work. "Underlying it all was the tremendous risk factor of would it be possible."
Boyer, for his part, later admitted the founders were green: "Bob and I were very naive about how we were going to do it."
Perkins did not reject the plan. He rewrote it. Rather than build anything, Genentech would subcontract the experiment to academic labs — Boyer's own at UCSF, and the City of Hope medical center in Los Angeles. No equipment, no building, no staff. The estimated $3 million start-up bill, as the documentary's narration has it, collapsed to roughly $250,000. Kleiner & Perkins committed $100,000 — a sliver of the firm's $8 million fund — in exchange for 20,000 shares of preferred stock at a May closing. Perkins was not sentimental about the odds. Asked about the risk of failure, he said: "Very high. I figured better than 50–50 we'd lose it." Then the other half of the sentence: "If it worked, the rewards would be obvious."
Genentech was incorporated on April 7, 1976. Swanson became president and treasurer, at a salary of $2,500 a month, with 25,000 shares. Perkins's line about that moment is the best one-sentence description of an early-stage company ever uttered:
— Tom Perkins, in the documentary Something VenturedIt was Bob Swanson and a checkbook, sitting in our office. That was Genentech.
Notice whose office. The firm that had shown Swanson the door a few months earlier was now, in a sense, his landlord, his banker, and his co-conspirator. It is tempting to read this as vindication, and the temptation should be resisted, at least a little. Perkins's genius was not in believing Swanson. It was in structuring a bet so that believing him was cheap. The company Swanson and Boyer imagined would have cost $3 million to find out it didn't work. The company Perkins imagined would cost a quarter of that. Both visions turned out to be necessary: Swanson's to make the thing worth doing, Perkins's to make it survivable.
The Real Thing
Insulin was the goal. The scientists quickly decided it was the wrong first step.
The reasoning was sensible. Rather than try immediately to coax a bacterium into synthesizing insulin — a protein with two chains that would have to be made and joined — they would start with somatostatin, a much smaller hormone, and prove the method on something tractable. If a bacterium could be made to produce a human protein at all, the rest would be engineering. If it couldn't, better to learn that fast and cheap.
Swanson hated it.
"If you are going to go for something," he said, "go for the real thing."
He gave in, grudgingly. This is one of the most revealing moments in the founding story, because it shows both what made him necessary and what made him difficult. His instinct was toward the market, toward the product people would pay for, toward the molecule that had a $100 million market waiting for it. Somatostatin had no such market. It was a demonstration, a proof of principle, a thing you did to persuade the world and yourself that the impossible was possible. Swanson wanted the prize. The scientists wanted the proof. The scientists won the argument, and the scientists were right, and Swanson — this is the part that matters — let them be right.
He set about formalizing research agreements with the University of California and with City of Hope, where the chemist Keiichi Itakura and the biologist Arthur Riggs, both of what is now the Beckman Research Institute, would synthesize the gene itself. In early 1977 he began a second round of financing and by February had raised about $850,000 — enough to fund the somatostatin work.
The plan was a relay race across the length of California. City of Hope, in Los Angeles, would build the gene from scratch. UCSF would splice it into bacteria and get the bacteria to read it. And according to the story as it is told in Something Ventured, when City of Hope finally succeeded after a very long time in making the gene, Bob Swanson carried it in his pocket to San Francisco, and Boyer put it into the bacterial host.
Again the reporting voice must intrude: this is a detail from a documentary's retelling, not from a lab notebook, and I would not stake my life on the specifics of which pocket or which flight. But consider what it means even as legend. The business partner — the man without a doctorate, the man in the suit — as courier. The chemist who left chemistry because he enjoyed people more than things, carrying, in the most literal sense, the thing.
By August 1977 the teams had made the first bacterium capable of synthesizing somatostatin. It was the first time anyone had expressed a human gene in bacteria to make a human protein. On December 2, 1977, Swanson and the scientists held a press conference.
Then, immediately, he pointed everyone back at the real thing.
A Warehouse by the Airport
The race for human insulin was a race. Two other scientific teams were already trying to do it, and Swanson's priorities for early 1978 were the three things that would let Genentech win: lab space of its own, corporate contracts, and more money.
The lab space came first. In February 1978, Swanson leased a 10,000-square-foot section of an airfreight warehouse in South San Francisco. Here is the second appearance of the airplanes in this story, and it will not be the last. The son of an Eastern Air Lines maintenance crew leader established the first purpose-built biotechnology laboratory in a building made for cargo, in the industrial flatlands near the airport, in a town whose company address would eventually become 1 DNA Way.
What went into the warehouse was more important than the warehouse. Two of the people who went in — David Goeddel and Arthur Levinson, both of whom would rise to the top of the field — later wrote that the move from a "paper" company subcontracting work to academic labs to a real company with its own bench space was "a critical juncture." Swanson now had to persuade young postdoctoral scientists to abandon the academic career path for a company that might not exist in two years, in a field that, as Goeddel and Levinson put it, "even leading molecular biologists thought... was neither commercially nor intellectually viable."
He did it, they wrote, "because he was able to generate his own conviction and excitement in others, as we know only too well — we were two of his earliest recruits."
Goeddel is worth a moment. He and Dennis Kleid joined the effort and became central to the insulin work; he would go on to clone a series of the company's most important early proteins and, much later, to found a company of his own, Tularik, where Swanson would serve as chairman. The arc rhymes with Swanson's own: a young man talked out of the expected path and into an unproven one, who eventually went on to start something himself and asked the man who had recruited him to sit at the head of the table.
Swanson also offered something that was not money and was, for scientists, more persuasive than money. Genentech researchers would be allowed to publish their findings in scientific journals — something most big drug companies were reluctant to permit. The only condition was sequence: publish after the patents were filed. It was a small clause that resolved a large contradiction. Academic science runs on priority and openness; commercial science runs on secrecy and property. Swanson's arrangement let both currencies circulate in the same building. The policy outlived him.
Money and contracts came together. In 1978 Swanson secured a relationship with Eli Lilly, the established insulin maker, under which Genentech would receive $50,000 a month to pursue human insulin. In August 1978, Genentech's scientists produced human insulin in bacteria. That same month, Swanson and his colleagues negotiated a multimillion-dollar contract with Lilly. Lilly would manufacture the product and shepherd it through the Food and Drug Administration; when the FDA approved Humulin in 1982, it became the first genetically engineered human therapeutic ever approved.
The structure of that deal — a small company with a technology, a large company with manufacturing, regulatory experience, and a sales force — became, in the words of Hughes's history, the eventual template for biotechnology start-ups. Every biotech partnership signed since carries some trace of the terms worked out in a warehouse by the airport in the summer of 1978.
The new revenue, and the press, bought room. By 1979 Genentech had projects underway on interferons, animal growth hormones, a hepatitis B vaccine, and the hormone thymosin. That year, according to MIT's account, it became the first major biotechnology company to show a profit. Swanson was thirty-one.
The Honeymoon Roadshow
In 1980 Swanson married Judy Church. In September of that year, the couple went to Europe. The trip was a honeymoon. It was also a sales tour to drum up interest among European investors ahead of Genentech's public offering.
One reads that detail and laughs, and then stops laughing. The honeymoon-as-roadshow is the sort of anecdote that circulates in business lore as evidence of devotion, and perhaps it was; it may also have been, for the bride, something closer to an omen. It was a marriage that lasted until his death and produced two daughters, Katie and Erica, whose soccer triumphs he would later recount to anyone who'd listen, including the president of MIT. Both things can be true. The record does not tell us what Judy Swanson thought of the itinerary, and I will not pretend it does.
Why go public at all? The company needed more money to continue, of course. But Swanson's stated reason was subtler: public interest in the technology was high, and he believed it should be capitalized on. He was reading a window. In 1980 genetic engineering was simultaneously terrifying and thrilling to the American public, and Genentech was its most visible face.
The offering took place on October 14, 1980. Genentech raised $35 million — at that moment, the largest IPO in American history. The stock soared on the first day of trading. Among the many people suddenly made rich, at least on paper, were the scientists Swanson had talked out of academia; Goeddel would later recall his own stake in interviews.
There is a cliché that founders lose control of their companies at the IPO. Swanson used the IPO to seize control of his company's direction. Before 1980, Genentech was in effect a high-end contract research shop — brilliant, profitable, and dependent on larger companies to turn its molecules into products. After 1980 Swanson pursued a different idea: Genentech as a self-sustaining biotechnology firm that would discover, manufacture, win approval for, and sell its own drugs. The insulin deal with Lilly had been the right first move. It was not, in his mind, a business model. It was a bridge.
Two years later the company would sign another such bridge, agreeing in 1982 to sell a 6.5 percent equity stake to Corning Glass Works and form a joint venture, Genencor, to develop industrial enzymes with recombinant DNA. The pattern was set: partner where partnership bought time, keep for yourself what would define you.
Hector, Pester, Hover, Noodge
In 1985, a young science writer researching a book called Invisible Frontiers had a lunch appointment with Bob Swanson, then president and CEO of Genentech. On his way south on U.S. 101, he got stuck in traffic — traffic, he later noted with some irony, created in part by the biotech revolution Swanson had helped launch. He arrived an hour and a half late, mortified.
Swanson "betrayed not the slightest irritation." They went to the company cafeteria, and Swanson recalled the early days at Genentech "with passion and humor."
Writing after Swanson's death, in an essay for MIT Technology Review titled "Strategic Patience," the writer reflected on that lunch. "I'm sure he was more peeved than he let on," he wrote, "but patience was the better part of valor." Then he immediately complicated his own thesis: "His supply of patience, let me hastily add, wasn't infinite. Having spoken with many of the scientists who formed the core of Genentech's original research staff, I know that Swanson could hector and pester and hover and noodge with the best of them."
Hector and pester and hover and noodge. Four verbs, escalating, the last one Yiddish, each a slightly different texture of impatience. That list is the most honest portrait of Swanson's management style in the record, and it sits right beside the story of the man who waited ninety minutes for a journalist without complaint. The two portraits do not contradict each other. They describe a man who had sorted the world into things that could wait and things that couldn't. A reporter could wait. A protein could not.
The New York Times obituary put it more bluntly: "Mr. Swanson was extremely hard-driving and demanding." It also noted, in the same paragraph, the culture he built around that drive — informal, academic in texture, known for its Friday afternoon themed parties, which everyone called "ho-ho's."
He made the work fun, his early scientists insisted. "It was Swanson who made all that hard work a lot of fun," Goeddel and Levinson wrote in Nature. "He led the celebration as each milestone passed." The milestones included insulin, growth hormone, interferon, factor VIII. And in 1987, when the FDA approved Activase — tissue plasminogen activator, a drug that dissolves the blood clots behind heart attacks — the celebration included, with permission from nearby San Francisco International Airport, a fireworks display large enough to temporarily disrupt air traffic.
Hold that image. We will come back to it.
Beneath the parties and the pestering was a standard. Goeddel and Levinson recalled that Swanson "reminded us why we were doing it — for the patients," and that he posed to employees what they called "the ultimate personal motivation": Would you put this drug into your own child's body?
It is a question that does several things at once. It is an ethical test. It is a quality-control protocol. It is a recruiting pitch — a reminder to postdocs who had given up tenure tracks that their work was not merely commercial. And it is, crucially, a question asked by a businessman, not a scientist. Swanson could not judge the chemistry. He could judge whether the people judging the chemistry would stake their own children on it.
"Swanson was a businessman," Goeddel and Levinson wrote, "but he never gave up hope in the science or the scientists during the early days. It was Swanson who was smart enough to hire the best people and give them an environment that allowed them to make a difference."
George Rathmann, who cofounded Amgen — Genentech's great rival — and was by 1999 chairman of the Icos Corporation, offered a competitor's tribute, which is the most credible kind: "Basically, he started the biotechnology industry in three or four key ways. The guy made the most extraordinary possible contribution to a whole industry."
What Growth Hormone Cost
The bridge Swanson built after the IPO led to human growth hormone.
He believed recombinant growth hormone had a large market in the United States and that it would be the molecule that carried Genentech from contract researcher to true pharmaceutical company. Like insulin, it was a target with an existing infrastructure of knowledge in academia and at the FDA — one reason, the Technology Review essayist noted, Swanson had "shrewdly chose[n]" both. Unlike insulin, Genentech would develop it almost entirely on its own.
On October 18, 1985, the FDA approved Genentech's human growth hormone, under the brand name Protropin, for children with growth hormone deficiency. Genentech had manufactured it, won federal approval for it, and would market it itself. It was the first time a biotechnology company had sold its own drug. Over roughly two decades, Protropin's sales exceeded $2 billion. Swanson's plan, conceived on a honeymoon roadshow, had worked.
The plan also sent bills, and they arrived late.
In 1990, the University of California, San Francisco — the institution in whose lab Swanson had first met Boyer, whose researchers had been Genentech's original subcontractors — sued Genentech for $400 million, alleging that Protropin had been developed using technology taken from the university and covered by a 1982 patent. Genentech maintained it had developed the drug independently. In July 1999 a jury found the university's patent valid but could not decide whether Protropin was based on UCSF research. In November 1999 Genentech agreed to pay $200 million to settle the nine-year-old dispute. The money was divided with an almost forensic precision: $30 million to the University of California general fund, $85 million to the three inventors and two collaborating scientists, $50 million toward a new teaching and research campus for UCSF, and $35 million to support university-wide research.
Earlier that same year, in April 1999, Genentech agreed to pay $50 million to settle federal criminal charges that it had marketed human growth hormone for unapproved uses from 1985 to 1994 — years that overlapped substantially with Swanson's tenure as chief executive. The Times obituary noted that Genentech's early days "were marked by some controversy over allegedly aggressive sales tactics."
Both settlements came in the last year of Swanson's life. Neither was a judgment on him personally. But it would be dishonest to write the founder's story without them, and they belong in the same paragraph as the child test, because they are the shadow cast by the same light. A company built by a man who could hector, pester, hover, and noodge — a company whose founder had resisted somatostatin because he wanted the real thing, now — was a company built for speed. Speed won the insulin race. Speed got Protropin to market. Speed, the settlements suggest, also had a price, paid on a schedule nobody chose.
By then Swanson had already made the biggest financial decision of his career. In 1990, Genentech agreed to sell 60 percent of its stock to Roche Holding, the Swiss drug giant, for $2.1 billion. Swanson stepped down as chief executive that February; Kirk Raab took over. Swanson became chairman, a role he held until 1996.
It is possible to read 1990 as the moment the dream of an independent biotech company ended — the start-up absorbed by the incumbent. That reading misses what actually survived. Genentech remained publicly traded and operationally its own creature. When Roche finally bought out the remaining shares in March 2009, for approximately $46.8 billion, its research arm was preserved as an independent center within Roche, still claiming, in its own literature, to "embrace the spirit of a 'startup.'" The scientists still publish. The airfreight warehouse company now anchors Roche's entire pharmaceutical operation in the United States. Swanson sold control to buy permanence. Whether that was the right trade depends on what you think a founder is for.
The First Company, Allegedly
A word about the claim that sits on every plaque.
Genentech is "widely considered the first biotechnology company." Cetus Corporation was founded in 1971, five years earlier. Genentech's own corporate overview, in a remarkable aside archived from its website, once admitted that "although the two confidently assert that it was the first biotech company, others clearly came before."
The Cetus problem is not a footnote. It is the hinge on which Swanson's career turned. He had tried, as a Kleiner & Perkins associate, to get Cetus to do exactly what Genentech would later do. Cetus said no. Swanson lost his job partly as a result. The first biotechnology company, in other words, had its chance to be the first recombinant DNA company and declined, and the associate who'd pitched them on it went off and built the company they declined to become.
So what does "first" mean? Genentech was not the first company to apply biology to industry. It was the first to make a human protein in a microbe, the first to clone human insulin, the first biotech to go public, the first to sell a drug of its own. More than any of those, it was the first to demonstrate — to venture capitalists, to entrepreneurs, to scientists — that a sustainable business could be built on genetic engineering. By the time of Swanson's death, Red Herring counted more than 1,000 biotechnology companies in the United States, and Goeddel and Levinson estimated the industry was worth more than $100 billion, with around 80 approved drugs benefiting millions of people.
"Almost no one," Red Herring wrote in its obituary, "will be able to say they founded a company that created an entire industry." The emphasis falls on created, not first. Swanson was not first. He was the one who made it a category. A field has to be shown to be possible before it is filled, and someone has to absorb the cost of the showing.
The Average Kid
After Genentech, Swanson formed a private investment firm, K&E Management, and backed start-ups. He became chairman of Tularik, the South San Francisco biotech company founded by David Goeddel — the postdoc he had talked into a warehouse in 1978. He served on visiting committees at MIT and for a time on the MIT Corporation, flying across the country to advise the institution that had once let a chemistry student slip into its business school. He was a trustee of the San Francisco Ballet. In 1986 he and Boyer had addressed student delegates at the Achievement Summit in Washington; the American Academy of Achievement awarded him its Golden Plate that year. Prizes accumulated: Entrepreneur of the Year honors in 1981 and 1983, Babson College's Distinguished Entrepreneur of the Year in 1993, membership in the Royal Swedish Academy of Engineering Sciences.
In 1998, the authors of 1,000 Years, 1,000 People ranked him number 612 among the men and women who had shaped the millennium — one of the few living people on the list — and called him the "entrepreneur who launched the biotechnology revolution."
He did not have long to enjoy it. Sometime around the end of 1998 he was diagnosed with glioblastoma, the most aggressive form of brain cancer.
The summer before he died, Charles Vest spent several hours with him. Swanson gave Vest a tie — he thought the president of MIT needed a better way to display the Institute's mascot, the beaver. The tie was covered with beavers beneath shining moons, and on close inspection the beavers were chewing down the Stanford pine tree. Vest said he wore it with pride. It is a perfect Swanson artifact: a gift, a joke, a rivalry, and a piece of institutional cheerleading, all embroidered into a strip of silk. Even dying, he was selling.
Robert Swanson died on December 6, 1999, at his home in Hillsborough, California, a few days after his fifty-second birthday. Genentech declined to comment, in deference to his family's wishes. MIT announced the death. Tularik was scheduled to go public that same week.
He was survived by his wife, Judy Church Swanson; his daughters, Katie and Erica; and his mother, Arline Swanson, of Foster City, California.
The honors kept arriving after him. That year he had been awarded the National Medal of Technology "for his foresight and leadership in recognizing the commercial promise of recombinant DNA technology"; Judy accepted it from President Clinton on her husband's behalf. In 2000 he and Boyer received the Biotechnology Heritage Award, his share posthumous. In 2006 he entered the Junior Achievement U.S. Business Hall of Fame. At Genentech's research complex, a life-sized statue commemorates the meeting with Boyer: the two men, fixed in place, at the start of something.
In 2009 — the year Roche paid $46.8 billion for the rest of the company that had begun as a $1,000 partnership — the FDA granted Genentech's antibody Avastin its fifth approval. The indication was glioblastoma multiforme.
The man who had asked his employees whether they would put a drug into their own child's body did not live to see whether any drug might be put into his. His company did not save him. It went on making things that might save someone else, which is the only thing a company like that can ever do, and the thing he had built it to do.
Holding Pattern
Go back, then, to 1987.
The FDA has just approved Activase, a drug that restores blood flow to a heart in the hours after an attack, a drug credited, in the obituaries a dozen years later, with saving countless lives. The company that made it began in a rented section of a building meant for freight. Its founder is thirty-nine years old, a man who once wanted to skip the small proof and go straight for the real thing, who learned to take the small proof, who learned to wait for the real thing, who never learned to wait for anything else.
Someone at Genentech has called San Francisco International Airport and asked for permission. Permission has been granted.
The father kept the planes in the air. That was the job: to work the shift, to check the wiring, to make sure the electrical systems held so the aircraft could take off on time and land on time and the schedule of other people's lives could go forward without interruption.
Over the flatlands of South San Francisco, the fireworks go up. Somewhere above the Bay, in the dark, the planes wait.
Part IIThe Playbook
What follows is an attempt to extract from Swanson's career the operating principles that made it work — and, in a few cases, the principles that made it expensive. Swanson did not leave a management treatise. He left a company, a template, a handful of quotations, and the testimony of the people he recruited, pestered, and celebrated. The principles below are inferred from what he did, not what he said he did, and they are offered with the caveat that the most important one — be in the right place when a science becomes a business — cannot be taught.
The principles
- 1.Let the failure fund the obsession.
- 2.Ask for ten minutes, then earn three hours.
- 3.Let someone else make your bet cheap.
- 4.Choose targets the world already understands.
- 5.Concede the proof, keep the prize.
- 6.Pay talent in the currency they actually value.
- 7.Rent the incumbent; don't become its vendor.
- 8.Raise into the public imagination.
- 9.Make impatience selective.
- 10.Install a test that outranks the org chart.
- 11.Celebrate loud enough to stop traffic.
- 12.Expect speed to send an invoice.
Principle 1
Let the failure fund the obsession
The heroic version of Swanson's story begins with vision. The accurate version begins with a severance. He was a venture associate who, by his own boss's account, worked deals "not very successfully," who spent his time trying to push a portfolio company into a field it didn't want to enter, and who was asked to leave because of it. The Antex bankruptcy that introduced him to Eugene Kleiner, and the Cetus rejection that ended his job at Kleiner's firm, were both failures. Both were also the precise mechanisms that put him, unemployed and unencumbered, in front of Herbert Boyer.
The non-obvious lesson is not "failure is good." It is that the thing you are failing because of may be the thing you should be doing. Swanson was penalized for an interest that the market had not yet priced. His firing was a signal that he was early — and early, with no employer to answer to, is a position of enormous leverage if you are willing to stop interviewing.
His decision rule was stated with almost comic modesty: "If I don't do this, I'm not going to like myself so much for not having given it a shot." That is not a vision statement. It is a regret-minimization calculation performed by someone who had already absorbed two failures and decided the third would at least be his own.
Tactic: When you're punished for an interest, ask whether you're wrong or merely early — and if early, treat the loss of your job as the loss of your last excuse.
Principle 2
Ask for ten minutes, then earn three hours
Swanson cold-called scientists from a list. Most declined. Boyer agreed to ten minutes, late on a Friday, motivated primarily by the possibility of lab funding. The ask was small enough to grant without thought. What Swanson did with it was the work.
Small asks lower the cost of a yes; they don't guarantee a conversion. Swanson converted because, as Boyer put it, he arrived with "the drive and the understanding." The chemistry degree he had abandoned as a career turned out to be decisive as a credential — it let him follow the technical nuances well enough that a scientist could talk to him as a peer rather than translate down to a financier. The ten minutes worked because he had prepared for three hours.
There is also a lesson in the target. Swanson did not call pharmaceutical executives. He called the people who had the capability and not yet the ambition to commercialize it. Boyer later said it was not his goal to start an industry, only to make sure the science became useful. Swanson supplied the ambition; Boyer supplied the capability. The ten-minute meeting was a matching of complementary deficits.
Tactic: Make your first ask so small it's rude to refuse, then show up with enough domain fluency that the other person forgets the clock.
Principle 3
Let someone else make your bet cheap
The founders' original plan called for roughly $3 million to build a facility and hire staff to test whether the science worked. Tom Perkins rewrote it so that Genentech would subcontract the experiments to UCSF and City of Hope, cutting the start-up cost to about $250,000. Kleiner & Perkins put in $100,000 — a small fraction of an $8 million fund — on odds Perkins privately rated "better than 50–50 we'd lose it."
Swanson's contribution here was the willingness to accept a restructuring that made his company look less like a company. For two years Genentech was, in Goeddel and Levinson's phrase, a "paper" corporation — no lab, no staff, just contracts, a checkbook, and Bob Swanson. Founders often resist this because the paper version feels small. But it compressed the most uncertain question — does this work at all? — into the cheapest possible container.
Perkins's philosophy was "put the risk up front and get rid of the risk as fast as you possibly can." The combination of Swanson's ambition and Perkins's risk discipline was more valuable than either alone.
Tactic: Before you build anything, find the cheapest structure that answers the question most likely to kill you — and let a skeptical investor design it.
Principle 4
Choose targets the world already understands
Swanson picked insulin, then human growth hormone. Neither choice was scientifically daring in its target, only in its method. Insulin's structure was already elucidated. There was a market of more than $100 million a year and a clear clinical reason — allergic reactions to pig insulin — for a human version. As the Technology Review essay observed, he "shrewdly chose two targets — insulin and human growth hormone — in part because they already enjoyed a large infrastructure of knowledge both in academia and at the Food and Drug Administration."
This is the less glamorous half of being a pioneer. When the platform is radical, the application should be legible. Every unknown you add to the product compounds the unknowns in the technology. Swanson's genius was in pairing a terrifying new method with molecules that regulators, doctors, and investors already understood.
◆
Why Insulin First
Swanson's target selection reduced risk on every axis except the one Genentech existed to solve.
| Risk dimension | Insulin in 1976 | Status |
|---|---|---|
| Scientific characterization | Structure already elucidated | Low |
| Clinical rationale | Pig insulin caused allergic reactions in many patients | Clear |
| Market size | World sales above $100 million and growing | Proven |
| Regulatory familiarity | Existing infrastructure of knowledge at the FDA | Moderate |
| Production method | Human gene expressed in bacteria — never done | Extreme |
Tactic: If your technology is genuinely new, point it at a problem everyone already agrees is worth solving, so the only open question is whether your method works.
Principle 5
Concede the proof, keep the prize
When the scientists proposed starting with somatostatin instead of insulin, Swanson objected: "If you are going to go for something, go for the real thing." He lost the argument, accepted the loss grudgingly, and then funded the detour fully — raising about $850,000 by February 1977 for a molecule with no commercial market.
The somatostatin result in August 1977, announced at the December 2 press conference, was the proof of concept that made everything after it financeable. Then Swanson immediately redirected the company toward insulin and won the race by August 1978.
The principle is not "defer to experts." It is subtler: let the people closest to the risk sequence the work, while you hold the destination fixed. Swanson never stopped wanting insulin. He simply accepted that the shortest path to it ran through something smaller. Founders who refuse intermediate proofs often lose the prize; founders who fall in love with intermediate proofs often never reach it.
Tactic: When your team proposes a smaller first milestone, agree to the sequence but restate the end goal out loud the day the milestone is hit.
Principle 6
Pay talent in the currency they actually value
Swanson's recruiting problem in 1978 was severe: persuade postdocs to abandon academic careers for a company in a field many leading biologists considered commercially and intellectually unviable. Equity helped, eventually. But the decisive offer was permission to publish — something most big drug companies would not allow — with one condition: patents first.
That single sequencing rule resolved the central conflict between academic and commercial science. Researchers kept the reputational currency that their careers ran on. The company kept the intellectual property its business ran on. The policy still stands at Genentech.
The general principle: the most valuable thing you can offer high-skill talent is often not money but the preservation of the status economy they came from. Swanson did not ask scientists to become businesspeople. He built a business in which they could remain scientists.
Tactic: Identify what your best recruits would lose by joining you, then design a rule that lets them keep it under conditions you can live with.
Principle 7
Rent the incumbent; don't become its vendor
The Eli Lilly deal — $50,000 a month, then a multimillion-dollar contract after insulin was cloned in August 1978 — gave Genentech revenue, validation, and a partner with the manufacturing and regulatory muscle to bring Humulin to approval in 1982. It became the template for the industry.
But Swanson treated it as a bridge, not a destination. After the 1980 IPO he pursued a self-sustaining company that would develop, manufacture, and sell its own drugs. Protropin, approved October 18, 1985, was the proof: the first biotech company to sell its own drug, and more than $2 billion in sales over two decades.
→
From Contract Lab to Drug Company
How Swanson sequenced dependence and independence.
1976
Genentech founded April 7 as a "paper" company subcontracting research to UCSF and City of Hope.
1977
Somatostatin expressed in bacteria; second round of roughly $850,000 raised by February.
1978
Airfreight warehouse leased in February; human insulin cloned in August; Eli Lilly contract signed.
1980
IPO on October 14 raises $35 million; strategy pivots toward a self-sustaining company.
1982
Humulin approved, manufactured and marketed by Lilly; Corning takes 6.5% stake to form Genencor.
1985
Protropin approved October 18 — Genentech's own drug, made and sold by Genentech.
1987
Activase approved for heart attacks.
1990
Roche buys 60% for $2.1 billion; Swanson becomes chairman.
Tactic: Take the incumbent's money for your first product, but use the time it buys to build the one capability that lets you sell your second product without them.
Principle 8
Raise into the public imagination
Swanson's stated reason for the 1980 IPO was not only that Genentech needed capital. It was that public interest in genetic engineering was high and should be capitalized on. He read the window correctly: the October 14 offering raised $35 million, the largest IPO in history to that point, and the stock soared on day one. He even turned his honeymoon into a European investor tour to widen the funnel.
The lesson is about timing relative to attention, not relative to readiness. Genentech had no approved drug of its own in 1980. What it had was a narrative — insulin made by bacteria — that the public found astonishing. Swanson sold into astonishment and used the proceeds to build the substance that would justify it.
This is dangerous advice in the wrong hands; the Technology Review essay's contempt for "dot-com dervishes who have yet to produce a product" is a fair warning. The difference is what the money is for. Swanson raised on hype to fund a decade-long march toward a real product. Most hype raises fund the hype.
Tactic: When the public is paying attention to your field, raise more than you need — and commit the surplus to the slowest, hardest thing on your roadmap.
Principle 9
Make impatience selective
The writer who arrived ninety minutes late for lunch found a man who "betrayed not the slightest irritation." The scientists who worked for him found a man who could "hector and pester and hover and noodge with the best of them." Both are accurate.
Swanson's impatience was aimed at the critical path — the insulin race, the next proof, the product the company needed "sooner rather than later." His patience was reserved for the larger arc: the conviction that recombinant DNA would "ultimately transform both science and pharmaceutical commerce. It was just a matter of time." The essayist named this combination "strategic patience."
Most leaders are uniformly patient or uniformly impatient, and both fail. Uniform patience lets races slip; uniform impatience burns out the people you need for the long game. Swanson sorted. A reporter's lateness was not on the critical path. A competing lab's progress on insulin was.
⏱
Where Swanson Spent His Patience
A rough sorting of what could wait and what couldn't.
| Situation | Response | Mode |
|---|---|---|
| Two rival teams racing to clone insulin | Moved fast on lab space, contracts, and funding | Impatient |
| Scientists' daily progress | Hectored, pestered, hovered, noodged | Impatient |
| A journalist 90 minutes late | No visible irritation; lunch with "passion and humor" | Patient |
| Transformation of pharma by recombinant DNA | "Just a matter of time" | Patient |
| Somatostatin detour | Resisted, then funded fully | Negotiated |
Tactic: List the three things on your critical path this quarter and allow yourself to be relentless only about those; be visibly generous about everything else.
Principle 10
Install a test that outranks the org chart
"Would you put this drug into your own child's body?" Swanson posed this question to his employees, and Goeddel and Levinson remembered it as "the ultimate personal motivation."
What makes the question powerful is that it cannot be answered by rank. A CEO cannot overrule it; a scientist cannot hide behind a protocol. It forces each person to translate abstract quality standards into a personal stake. And it was asked by a non-scientist who could not himself evaluate the molecules — which made it a way of borrowing his researchers' judgment and binding it to their consciences.
Every organization needs a small number of questions that override hierarchy. Swanson's was simple, emotional, and impossible to game.
Tactic: Write one question about your product that any employee can ask of any decision, and make sure leadership is subject to it first.
Principle 11
Celebrate loud enough to stop traffic
The ho-ho's — Friday afternoon themed parties — and the 1987 Activase fireworks, which briefly disrupted air traffic at SFO with the airport's permission, were not decoration. Goeddel and Levinson made the point directly: Swanson "made all that hard work a lot of fun. He led the celebration as each milestone passed."
In a field where products take years and most experiments fail, milestones are the only rewards that arrive on a human timescale. Celebrating them extravagantly does two things: it marks progress for people who might otherwise only see the remaining distance, and it signals to recruits — who are, after all, choosing between you and a tenure track — that this is a place where achievement is witnessed. The informal culture, the Times noted, was part of how Genentech competed for academic talent.
Tactic: For every hard milestone, budget a celebration proportionate to the effort, not the revenue — and let the founder lead it personally.
Principle 12
Expect speed to send an invoice
The same drive that won the insulin race produced, in Swanson's last year of life, two large bills. In April 1999 Genentech paid $50 million to settle federal criminal charges of marketing human growth hormone for unapproved uses from 1985 to 1994. In November 1999 it paid $200 million to UCSF to settle a nine-year patent dispute over the origins of Protropin — the very drug that had proved Swanson's independence strategy.
Neither settlement erases the achievement. Both show that a culture optimized for speed accrues liabilities that come due long after the decisions that created them, often under different leadership. The UCSF case is especially instructive: the institution where Genentech was born became its most expensive adversary, over the boundary between academic and commercial work that Swanson had otherwise managed so deftly.
The principle for operators is not "slow down." It is that a hard-driving culture needs an equally deliberate counterweight — in compliance, in IP hygiene, in relationships with the institutions you came from — because the bill for speed is paid in a currency and on a timeline you do not control.
Tactic: For every system you build to move faster, assign someone senior whose explicit job is to imagine the lawsuit ten years out — and give them the authority to slow one thing down.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIQuotes and Maxims
In their words
— Robert Swanson, 1996 interview, on leaving a career in chemistryOne of the things I discovered was that I enjoyed people more than things. So I said, 'Gee, this probably isn't going to be what I'd want to do all my life.'
— Robert Swanson, objecting to starting with somatostatin instead of insulinIf you are going to go for something, go for the real thing.
— Robert Swanson, as recalled by David Goeddel and Arthur Levinson in NatureWould you put this drug into your own child's body?
— Robert Swanson, as recalled by MIT President Charles VestWhat's going to happen to the average kids like me?
— George Rathmann, cofounder of AmgenBasically, he started the biotechnology industry in three or four key ways. The guy made the most extraordinary possible contribution to a whole industry.
For the definitive account of the company's founding, drawn from extensive interviews including Swanson's own oral history, see Sally Smith Hughes's Genentech: The Beginnings of Biotech.
Maxims
- Being early can look like being wrong. Swanson was let go for an interest the market hadn't priced yet; the firing was a signal, not a verdict.
- Small asks open large doors. Ten minutes is easy to grant; what you bring into those minutes decides whether they become three hours.
- Let a skeptic price your bet. Perkins turned a $3 million science project into a roughly $250,000 experiment, and that made the dream survivable.
- Radical methods need legible targets. Point a new technology at a problem regulators, doctors, and markets already understand.
- Take the small proof, keep naming the big prize. Somatostatin bought the credibility that insulin needed.
- Pay people in their native currency. Letting scientists publish, after patents, recruited talent that money alone could not.
- Partner to survive, build to own. The Lilly deal was a bridge; Protropin was the destination.
- Raise when the world is watching, spend on what takes longest. Hype is a financing tool, not a strategy.
- Spend impatience only on the critical path. Hector the race; be gracious about everything else.
- Speed borrows against the future. The bills for a fast culture arrive late, and someone always pays them.
