Part IThe Story
"I am not someone who gets stuck in the rut," Sandy Lerner said. A few minutes later she drove into one.
This was in 2014, on Ayrshire Farm in Upperville, Virginia. A German journalist named Sabine Muscat was in the passenger seat, and the photographer Kate Warren was somewhere behind her with a camera. They had been talking about range, about the improbable line that ran from computer programming to novel-writing to organic agriculture, and Lerner, at 59, had offered the remark as a summary of herself. Then the four-wheel drive left the road and settled into a field. The wheels spun. The reporter, the photographer, and the woman who had co-founded the company that sold the routers of the early internet sat there for a while, stuck in exactly what she had said she never got stuck in.
She did not call anyone. She shifted gears, let the wheels turn, and steered backward an inch at a time until the tires found purchase. Within minutes they were back on the track, driving past pastures of heritage-breed horses and pigs toward the farmhouse and the barn across from it that she used as a silversmith's studio. Inside there was Earl Grey tea, copper pots, and a cat asleep in the kitchen. Muscat wrote that the room felt like a film set for a Jane Austen novel, and it was meant to. Lerner had by then read Pride and Prejudice more than a hundred times. She had written a sequel to it.
A router is a machine for choosing paths. It looks at where a packet has to go and decides how to get it there, across networks that were never designed to talk to one another. Lerner spent the 1980s helping build and sell those machines, and the decades after that turning down whatever road someone else had laid for her: the venture capitalists' road, the cosmetics industry's road, agribusiness's road, the Virginia hunt country's road. The wheels spun a great deal. She was fired from her own company 193 days after it went public, and she has never fully stopped being angry about it, or stopped talking about it with the bluntness of someone who learned an expensive lesson and won't let anyone pay for it twice.
The scene in the field is small. It holds most of what follows, though: the insistence, the comedy, and the stubbornness the British call bloody-mindedness, a phrase she is fond of. What gets her out of a rut is backing up slowly and patiently. Running at it does not work.
By the Numbers
A Life in Routes
$109KCisco's total sales in its first eight months, through July 31, 1985
$2.6MWhat Don Valentine paid for '30-odd percent' of Cisco, by Lerner's account
193Days between Cisco's IPO (Feb. 16, 1990) and Lerner's firing (Aug. 28, 1990)
$170MWhat Lerner and Bosack's Cisco stock fetched when they sold it all
70%Share of their Cisco proceeds the couple gave to charity
$500B+Cisco's market capitalization in March 2000, when it briefly passed Microsoft
8,000+Early women's books Lerner collected and donated to Chawton House Library
Two Aunts, One Steer
Her parents divorced when she was four. She grew up between two aunts who lived very different lives. One had a cattle ranch at Clipper Gap in the California Sierras. The other lived in Beverly Hills. Profiles of Lerner tend to treat this as a quaint detail. It is closer to a key. Nearly everything she has done since can be sorted under one aunt or the other, and the most interesting things she has done, such as an organic farm with a 42-room Edwardian mansion or a cosmetics company that sold eyeshadow called Gangrene, needed both.
The ranch came first in her account of herself. At nine she bought a steer, kept it two years, and sold it at a profit. With the proceeds she bought two more. She spent nine years in 4-H and was by one account an all-star. She owned a Welsh cob horse named Blackjack. By the time she reached college she owned a registered herd. One account puts it at 30 head, and a later retelling inflated it to 200, which is what tends to happen to herds in the telling. She sold the herd in stages to pay her tuition. The detail matters. Her first capital came from cattle she raised herself, and she owed nothing to anyone for it. That remains her ideal of how a venture should be financed.
She went through California State University, Chico, in two years and finished in 1975 with a degree in political science. In a 2010 profile, Thomas Heath of The Washington Post described the course of study as "comparative communist theory with a minor in Marxist economics." The future capitalist was, as Heath put it, a "onetime socialist." She seems never to have treated the label as an embarrassment. She treats it as one more phase, like the punk eyeshadow and the jousting.
She went on to Claremont Graduate School outside Los Angeles to study comparative politics. In 1975 she wandered into the school's new computer lab and kept coming back until they made her its manager. She began using the machines for quantitative analysis in social science, which at the time hardly anyone did. "People left you alone because they thought you were weird," she told Genevieve Leone of the San Jose Business Journal, "and nobody knew anything about what you were doing."
That sentence describes a kind of freedom. The lab was a place where nobody understood the work well enough to interfere with it. She would spend the next fifteen years in places like that, and when the people who did understand finally showed up, she would leave.
She took a master's in econometrics from Claremont in 1977 and moved north to Stanford for a second master's, in statistics and computer science. She described the admissions committee's reasoning without sentiment: "Stanford was dying to get graduate students back then who could bail out the faculty who didn't know anything about computers. If you had all these hacker-type graduate students, you'd moved into the computer age. So they let me in."
Leonard Bosack was born in Pennsylvania in 1952 to a Polish Catholic family. He finished La Salle College High School in 1969 and the University of Pennsylvania's engineering school in 1973. He went to Digital Equipment Corporation as a hardware engineer, the sort of job that teaches you what machines can actually do as opposed to what the brochures claim. In 1979 he came to Stanford to study computer science. Muscat would later describe Lerner's world as a Jane Austen set. Bosack's was the server room. He was, in the phrase of one networking history, a "computer nerd," and he seems never to have wanted to be anything else. Decades after Cisco he would be running a small optical-networking company called XKL and talking about the physics of fiber amplifiers, which he had used to move data roughly the distance from Chicago to New York, 1,231 kilometers, in 6.071 milliseconds.
The two met in 1977 while sharing time on Stanford's minicomputers. They married in 1980. By then each ran a computer facility, Lerner at the Graduate School of Business and Bosack at the Computer Science Department. Their buildings were not far apart, but their machines were effectively in different countries, with different operating systems and different protocols, and none of them could talk to the others.
There is a legend about this, and it has been repeated so often it has hardened into a founding myth. In the legend, two lovers on opposite sides of a campus build the first router so they can send each other letters. Cisco's own Wikipedia entry and several of the network histories call it untrue. It survives because it is better than the truth. The truth was mostly about institutional plumbing: thousands of incompatible machines on a 16-square-mile campus whose users wanted them to behave as a single system.
Other people had been at the problem for years. By June 1980 Bill Yeager, a research engineer at the Medical School, had a working router. Stanford people would come to call his box the Blue Box. Its software was Yeager's, and its circuitry came from Andy Bechtolsheim, a graduate student so involved in Stanford hardware projects of the period that one historian calls him simply "ubiquitous." He would go on to co-found Sun Microsystems. Yeager's design scaled well. One account calls it "a key to Cisco's early success," and Yeager would later say he had been denied permission to sell it himself. In this story he plays a role Lerner would come to understand very well: the person who built something and watched others carry it away.
Yeager's router did not yet connect to the Ethernet network at the center of the Stanford University Network. In 1982, by one detailed history, Bosack, Lerner, and several other engineers "worked without authorization" to build an interface between Yeager's router and Ethernet. Then they pulled the coaxial cable across campus themselves. The project was a skunk works, and it worked.
The coaxial cable comes back at the end of this piece. For now it is enough to picture it, heavy and black, pulled through conduits and between buildings by people who had not asked anyone's permission. Stanford's administration would not forget it. Neither did the people who got the network.
They began building more routers in their living room, testing them on Stanford's network and on the IMPs that connected to the ARPANET. Word spread through the email of the day to other universities and research centers. Orders came in. The couple went to Stanford's Office of Technology Licensing to ask whether the university would build and sell the boxes or let them do it. The office, according to that history, "could not offer a solution that would take less than years."
The historian James Pelkey describes the Bosack and Lerner of that period as "happy academics willing to give what they had or knew away, resistant to even marking-up their production costs to make a profit." It is an odd thing to say about two people who would end up with $170 million. It is also, by every available account, true. They left Stanford partly because the university would not let them give the routers away fast enough.
A Company in the Living Room
On December 10, 1984, Bosack and Lerner incorporated cisco Systems. They spelled it in lowercase, and the company's early engineers insisted on the lowercase for years. The name was the tail end of San Francisco, and the logo, which survives in stylized form, showed the towers of the Golden Gate Bridge. A suspension bridge is a sensible emblem for a company that joined incompatible networks, and so is the lowercase, which suggests people who did not much want to look like a corporation.
In practice, according to the historian, "little changed other than that they could now take orders from any organization." In the roughly eight months to July 31, 1985, sales came to $109,000. The couple mortgaged their house as far as it would go, ran their credit cards to the limit, and borrowed from friends. Lerner took a day job running research computing for a Schlumberger laboratory. Kirk Lougheed remembers it as the Fairchild AI Lab, Fairchild then being a Schlumberger company. She worked on Cisco in the evenings and on weekends.
Lougheed worked as a systems administrator at Stanford, and he became Cisco's employee number four. His first badge read 004, and thirty-four years later his directory ID was still 00105. In a reminiscence the company published in 2020, he said that before he quit Stanford he had read "all the books on 'here's how people have started companies in Silicon Valley.'" None of them, presumably, described his first day. He showed up in July 1986 at a house in Atherton owned by Bosack and Lerner. One spare bedroom belonged to the administrator, a woman he remembered only as Cecilia. The other was the hardware lab. He sat at a table with a large HP workstation monitor and started writing the software that became Cisco IOS.
"In the next couple of weeks," he said, "we built and tested the first dozen routers in the living room. Then we made a couple of trips down to UPS to ship them to our first customers — some universities and HP." He added, a little dryly: "It's not exactly a standard first day at work."
Nobody drew a salary for the first two or three months. Lougheed lent the company money. The team was filled out by Greg Satz, a programmer, and Richard Troiano, who handled sales, along with what Lougheed called "a ragtag bunch of people" for final assembly. Sometimes they couldn't afford parts for prototypes. Ohio State University was the first paying customer. According to the IEEE Computer Society's account, the first month brought contracts worth more than $200,000.
Lougheed made a claim that Lerner has repeated many times, usually with heat: "We were actually profitable from the beginning."
Stanford, meanwhile, had not finished with them. On July 11, 1986, Bosack and Lougheed were forced to resign. The university considered filing criminal complaints against Cisco and its founders over the software, the hardware designs, and other intellectual property the company had taken. In 1987 the two sides settled the way institutions usually do, and Stanford licensed the router software and two computer boards to the company. Fifteen years later the dispute over credit flared up again. In December 2001 the San Jose Mercury News reported that a Stanford web page credited only Bosack and Lerner with the device, while a Cisco spokeswoman, Jeanette Gibson, called it a group effort. The Bosack entry concludes, in a sentence that could be engraved over the door of every startup: "Due to the nature of the collaboration, it is unable to be determined who did what during the process."
Public Protocols, Private Box
Lerner's own explanation of what Cisco sold, given years later on Guy Raz's How I Built This, comes down to a paradox. The communication protocols were in the public domain. They had been worked out collaboratively, largely at places like Stanford, by people who expected no payment. Cisco took that shared work and made a box people wanted to buy.
"In Cisco, we made a lot of money on public protocol," she told a CNBC reporter at a women founders' conference in 2012. The company's advantage was not the idea of routing. It was having made the idea reliable: something you could ship by UPS to a university or to HP, that kept working after it arrived, and that someone would help you with when it stopped. Business historians of networking have argued that companies like 3Com and Cisco succeeded by meeting rapidly rising demand from users in large organizations, "who were connecting computers into networks and networks into internetworks." The government built the protocols and the academics built the prototypes. Startups taught customers why they needed the product.
This is where Lerner's part in the story is both easiest to underrate and easiest to exaggerate. Bosack was the router engineer. Lerner, by the IEEE's account, "headed up all departments except engineering," and that covered nearly everything that turns a laboratory capability into a business. In the period before her firing she was vice president for customer service; Heath's profile calls it "a customer advocacy group." One later tribute calls her Cisco's "operational architect" and gives her credit for support systems and deployment practices. Its prose is so laudatory that it reads more like a nomination form than an account, and on that point the tribute should be read with some skepticism. The core of the claim holds up, though. Somebody had to decide that a box from a living room in Atherton would be supported like infrastructure. Lerner was that person.
Her later advice about intellectual property is unusually unsentimental for someone who profited from an engineering company. "There is no good idea that nobody can copy if you are playing with all public tools," she said in 2012. "Even if you have that technical IP, it will only buy you time before someone reverse engineers it. But what you can do is buy yourself time." She means it literally. A patent or a proprietary design is a head start, not a moat. After the head start, she said, "You just have to stay in the rat race. Be better, faster, cheaper."
A woman who learned to give her work away at Stanford, and then learned at Cisco how much that work was worth to other people, speaks about intellectual property the way a cattle rancher speaks about weather. It is real and it matters. You cannot count on it.
The Bozo No-No
Cisco looked for venture money for three years, by Lerner's count. She says seventy-odd firms turned it down, though the figure in the record is "three years," and she tends to round up when she is angry. At last, in 1987, Sequoia Capital's Don Valentine agreed to invest. Sequoia's website today lists both founders with a single line apiece: "Partnered In 1987." That is about as much feeling as the relationship would ever show.
She has told the story of the deal so many times, in nearly the same words, that it reads like scripture. The fullest version appeared in 2013 under the headline "Investors Aren't Friends":
"After funding the company for three years by mortgaging everything we owned and putting everything on credit cards, we made an absolutely bozo no-no. We decided to take money from a VC, Don Valentine. He got 30-odd percent of the company for $2.6 million. Len and I were very naïve. We used Don's lawyer and agreed to a four-year vesting agreement. We would get 90 percent of the founder's stock after four years. But we didn't get an employment contract."
Most founders who tell this kind of story hide the self-blame under the grievance. Lerner puts it first. "Bozo no-no" is the phrase of someone who has replayed the scene many times and decided the joke is on her. The lawyer was Valentine's. Nobody thought to ask for an employment contract, because the idea that the investor might one day want the founders gone had not occurred to anyone in the house in Atherton.
"My family had a small business," she said. "I always thought that if someone invested in your business, that meant he or she believed in it."
That is a cattle rancher's logic. If you put money into a steer, you want the steer to do well, and the steer and the money share one fate. Lerner had never met an investor who could want the company to do well and want the founders out in the same breath, or for whom those two wishes were really one. "I did not realize," she said, "he had decoupled the success of the company from that of the founders."
Lougheed's version is a fair footnote: "We never used the money that Sequoia provided to us. But what we had were some people who knew how companies work." The IEEE gives the figure as $2.5 million and says that "in late 1987, with the help of an infusion of $2.5 million in venture capital, Cisco began to thrive." Whether $2.5 million or $2.6 million, and whether spent or left in the bank, the investment's lasting effect was on governance, not on cash. A cash-flow-positive company gave away its control in exchange for what it was told was experience.
The experience showed up in the form of executives. Bill Graves was the first CEO, from 1987 to 1988. In 1988 John Morgridge took over. He would run the company through the IPO and hand it in 1995 to John Chambers, who would take it to half a trillion dollars. Morgridge left little colorful public record behind. One detail does survive. In 2012 the Stanford Graduate School of Business invited him to give its graduation address. That was the school whose computers Sandy Lerner had once run, and whose network she had once tied to the rest of Stanford with cable she had no permission to pull.
— Sandy Lerner, 'Investors Aren't Friends' (2013)At Cisco, I made every decision based on what was good for the company, and that pretty much ruined my marriage and my health. Len and I believed a company and its founder are the same. They absolutely are not.
August 28, 1990
Cisco went public on February 16, 1990, with a market capitalization of $224 million. That was the moment Lougheed would remember with relief: "It's like, 'This worked!'"
On August 28, 193 days later, Sandy Lerner was fired.
No source settles exactly who did it or exactly why. The records describe "Cisco's management" and "the professional managers the firm's venture capitalists brought in." Later retellings describe a clash between two unconventional, hands-on founders and a board that wanted a conventional corporation. Lerner's own explanation, given in the late 1990s, is structural and not personal: "The venture people were into building a bigger business and we were into building the best networking computer that you could build. They're not the same." She described those years as "tense" and drew a distinction between herself and her husband that she rarely draws so plainly: "I think I was very frustrated but (my husband) was kind of quieter about it."
When Bosack heard, he resigned. Every account says so, and every account uses the same word, solidarity or protest. It is about the most romantic act on record in the history of enterprise networking, and the record presents it in a single sentence.
Then they sold everything. Most sources give the figure as $170 million. Heath's 2010 profile says $200 million and adds that the couple had together owned a third of the company. The difference of $30 million is the sort of rounding error that only shows up in stories about people who got rich, and Lerner herself seems hardly to have noticed it. The money went out in a single block. Later she would describe it, with an eye-roll you can almost hear, as being "in California somewhere in a big pile and I pay someone to keep sweeping it up and doing things to it."
They divorced in the early 1990s.
In 2011 she agreed to an interview for Something Ventured, a documentary about the venture capitalists who built Silicon Valley. The film's producer later told an audience how hard she had been to persuade. One viewer wrote that the most striking thing in the interview was "the utter acerbity she still harbors about the ousting," two decades later. In it she offers a line that is either a precise historical claim or a deliberately compressed one: "I got fired by the same guy who fired Steve Jobs." It is not clear which. In the version of events she lives with, she belongs in the same category as Jobs, the founder thrown out of the company she built, and she has no intention of being quiet about it.
In March 2000 Cisco was briefly the most valuable company in the world, at more than $500 billion. The shares Lerner and Bosack had sold for $170 million would have been worth a sum it is pointless to calculate. On the How I Built This episode, a stretch of the conversation is listed in the show notes as "Missing out on a lot of money." It comes a few minutes after "Keeping a partnership with Len" and a few minutes before "Luck vs. hard work." She had arranged her feelings about it in that order: Len first, then the money, then a question that has no answer.
"When I was fired," she said, "it was devastating. I spent years crawling out from that."
That sentence says crawling, and in the field in Virginia she reversed. In neither case did she run.
Does Pink Make You Puke?
The way out of that started with a remark from one of the aunts.
Now that her niece was rich, the aunt suggested, perhaps she might start dressing like a well-to-do lady. She might wear makeup.
Lerner went to the cosmetics counters and found what she later described as a wall of pink and red, hundreds of variations on one idea of what a woman should look like. "I hate pink," she told an interviewer. "My skin is not pink. There was nothing about me that was pink." On another occasion she was more direct: "Fundamentally, I was just pissed off that cosmetic companies were telling women they had to look like Barbie."
So she started a cosmetics company. She put part of her Cisco proceeds into a venture partnership called & Capital Partners, and one of its investments, which by most accounts she co-founded, was Urban Decay. It was founded in the mid-1990s; one reference puts it at 1995. The colors were named for things nobody else would put in a lipstick: Smog, Mildew, Rust, Oil Slick, Acid Rain, Roach, Rat, Storm Drain, Polyester Bride, Gash, Gangrene. The first ad campaign asked, in large type, Does Pink Make You Puke? The second was headlined Burn, Barbie, Burn and was pulled when Mattel threatened to sue. Forbes used the first slogan as the headline of a 1997 profile. For that profile, according to several later accounts, Lerner posed naked on horseback, in the manner of Lady Godiva.
The Washington Post's Linton Weeks met her in 1998 and found purple dye in her long black hair, storm-cloud-blue polish on her nails, and "bruise-blue shadow above and below the eyes." She looked, he wrote, "a mite like Morticia Addams." She took it as a compliment. She loved The Addams Family, though only the television version and not the "drecky" film, and another account says she valued the Addamses because they taught tolerance. They were a household of oddballs who were perfectly happy in their big house and never wondered whether the neighbors approved.
"A number of people have tried to get me to grow up," she told James Sterngold of The New York Times.
The beauty business reported Urban Decay's success with mild surprise. Three years in, the line was in Nordstrom, Neiman Marcus, Bloomingdale's, and Harvey Nichols in London, and it was expected to gross $12 million a year. The products were never tested on animals. In February 2000 the French luxury conglomerate LVMH bought the company, the same month Cisco was heading toward half a trillion dollars. In December 2002 LVMH spun it off again.
Her analysis of why it sold is clear-eyed, and she does not flatter herself. Cosmetics, she said, is a bad business: "There is a relatively small kit of tools and everyone is playing with the same one." No intellectual property can be defended and no technical advantage holds. What Urban Decay had, after six years, was "brand cred," a reputation for being "really good makeup." LVMH paid for the reputation. "I am very lucky," she said.
In 2012 she repeated that phrase, "I am very lucky," in the same conversation where she said Cisco's founders had been "taken to the cleaners." She does not appear to see any contradiction.
The two companies look like opposites, one making routers and the other making eyeshadow called Gangrene. They have more in common than that. Each began with a group of insiders tired of an industry handing them a single option. Each turned a widely available set of tools into a product people would pay for. Each was eventually acquired by someone else. Lerner was rich enough the second time to find that funny.
— Sandy Lerner, on Urban DecayFundamentally, I was just pissed off that cosmetic companies were telling women they had to look like Barbie.
The Rich and Powerful Animal
Lerner and Bosack had largely given the money away before Urban Decay sold its first tube of Roach. Separated but on friendly terms, they put about 70 percent of their Cisco proceeds into charitable vehicles. The main one was the Leonard X. Bosack and Bette M. Kruger Charitable Foundation, reportedly endowed with $34 million. Both remained directors decades later. One reference work called her "the first female philanthropist to emerge from the Silicon Valley boom era." She did not behave like the ones who followed.
The grants were not the ones a newly rich founder makes to buy social standing. They went to Harvard and Stanford, to the SETI project, to the Center for Conservation Biology at the University of Washington, and to the Center for Wildlife Conservation. About a third went to animals: Born Free USA, the International Fund for Animal Welfare, the Austin Zoo, the Association of Veterinarians for Animal Rights, the African Elephant Conservation Trust, SPAY/USA. She did not stop at writing checks. She went back to building networks. She helped create Petlink, a database of lost cats and dogs and of animals available for adoption, and the IEEE notes her work on "creating and improving animal shelter internetworking." She took the same problem she had worked on at Stanford, scattered systems that could not find each other, and applied it to strays.
"I have yet to meet a rich and powerful animal," she said in the late 1990s. "You know, they don't vote and they don't talk. They're such a part of our life. We use them. We eat them. We make money from them. We exploit them. Somebody needs to watch out for them."
The list in the middle of that quote, we use them, we eat them, we make money from them, sounds a lot like how she describes what happened to her at Cisco. It would be too neat to call her animal advocacy a sublimated grievance. She raised a steer at nine and was a 4-H all-star long before there was a Sequoia term sheet. Still, a woman who learned in her mid-thirties that the people who financed her could decide she was no longer needed seems to have found that experience made the animals' position easier to see. She has watched a creature be worth a great deal to everyone except itself.
About money she was equally blunt. "I made a very conscious decision not to let it ruin my life," she said. "I think money is a good tool and I also think it is a drug."
And on whether she counted as eccentric: "I am now that I'm rich. I used to just be weird."
That is the Claremont computer lab again, twenty years later, with money in place of being ignored. The money did not change how she behaved. It changed how other people described it.
All the Old Neglect of Prospect
In 1996 she bought Ayrshire Farm in Upperville, Virginia. Wikipedia gives its size as 600 acres, a farm magazine as 793, and several reporters as 800, the sort of disagreement that happens when the subject keeps adding easements and neighbors keep estimating. On it stood a fieldstone Edwardian mansion from around 1912. Depending on who was counting it had 40 or 42 rooms, 18 bedrooms, and nine baths, and a furnace that burned sixteen gallons of oil an hour. She remembered the land as "overgrown." A farm magazine borrowed a phrase from Austen's Emma to describe its condition: "all the old neglect of prospect."
Why Virginia? Her answers vary in seriousness. "I like the font on the license plate," she told the Piedmont Virginian. "I am a jouster and the state sport of Maryland is jousting." Maryland is a different state. She did not seem to mind. Elsewhere she said she took up jousting in period costume because it offered "a chance to spear cabbages." When pressed, she gave the real reason: "Mostly because there is water here and farms that are large enough to farm but not large enough to be interesting to the agribusiness cartel."
She rejected the obvious way of describing the move. Everyone assumed the big change had been from Silicon Valley to farming. "Actually," she said, "the big transition was from farming to Cisco, or at least from our small farm in the Sierras to graduate school in Los Angeles and then Stanford. Moving back to a small town and back to farming was the easy part."
Upperville did not find it easy. The town's main annual event had been a horse show with a $100,000 Jumper Classic. Lerner arrived with purple hair, a reputation as a nude Godiva, and a refusal to let the local hunt ride over her land, and the town's established families reacted. Hate mail came. Bumper stickers reading DON'T MIDDLEBURG UPPERVILLE appeared on BMWs. When she tried to open an English-style pub, because the only lunch nearby was a country store she said made "terrible ham sandwiches," neighbors held it up for a long time. "The most vocal people have very little going on in their lives," she said. "The danger of the small town is that the mind sort of takes on the size of the town."
She won, slowly. The pub opened as Hunter's Head Tavern in the restored Carr House and served meat from Ayrshire. Under her management the farm became the first in Virginia to be Certified Humane and Certified Organic, and later Certified Predator-Friendly. Its veal operation, restaurant, and butcher shop were described as the first in the United States to be Certified Humane. She founded the Virginia Organic Producers' and Consumers' Association. She held an annual "Beastie Bazaar" for abandoned animals at the manor. She brought a Shire draft horse from California named Gronant Aristocrat, called Izzy, who stood eighteen hands. She translated a 1922 German manual on carriage driving, Benno von Achenbach's On Hitching and Driving, apparently because nobody else had and she wanted to read it. Visitors could book a ride in a restored carriage for $650 an hour.
By 2010 Ayrshire was a $7-million-a-year operation with 100 employees, from bartenders to a soils engineer, organized into seven business units. It raised 800 head of cattle, hundreds of pigs, hundreds of veal calves, thousands of chickens, and 3,000 heritage-breed turkeys for Thanksgiving. A 22-pound bird cost $230, and Lerner said she still lost money on it. Her customers included the Inn at Little Washington and Restaurant Nora in the District. When Heath, a Butterball man, admitted to her that he didn't care much about eating organic, she told him: "If you are not eating heritage-breed [turkeys], you are eating a hybridized, genetically engineered freak. You are what you eat."
The farm lost about $1 million that year, mostly on the cattle.
"I can sustain the loss indefinitely," she said. "But I don't want to."
That remark matters more than it first seems to. The idealist could absorb the loss. The rancher, who sold her first steer at a profit when she was eleven, did not want to. Ayrshire was meant to show that humane farming could pay for itself, and that if a market existed for heritage breeds, the market would keep the breeds alive. A charity cannot prove that. Only a business can.
She drove a Chevy Volt and, before long trips, plugged a small refrigerator in the back seat into the cigarette outlet and stocked it with local organic food. She had been promoting that kind of food for twenty years by then.
In 2004 she put a conservation easement on Ayrshire, which also protected the view of the Civil War battlefield at Upperville. She helped do the same for Mountain Gap Farm, Jubilation Farm, and others, more than 2,000 acres of Loudoun County in all. In 2022 the Loudoun Laurels Foundation named her its Laureate. It praised her for "selflessness, integrity, creativity, wisdom and courage." The hunt club's view was not recorded.
In 2013 she told The Washington Post she was ready to say goodbye to the farm. "I think everyone who makes money goes through a 'house phase,'" she said, "and I loved restoring the manor house, but it's a lot of upkeep." In time she sold it. Nearly 400 pieces of her collection went to auction at The Potomack Company: English furniture, Spode chocolate cups, a Manet etching of a cat, a cast-iron garden bench attributed to Coalbrookdale, sixteenth-century horse armor. The auctioneer said what struck her was that the objects had been used. "It was a lived-in collection." The armor had presumably been worn by a jouster.
Second Impressions
In 1992, through the foundation she shared with Bosack, Lerner took a 125-year lease on Chawton House in Hampshire, England. It had belonged to Edward Austen Knight, Jane Austen's brother. His living there is what allowed Jane, her mother, and her sister Cassandra to settle in Chawton Cottage down the lane, where Jane wrote or revised all of her novels for publication. When Lerner found it, the great house was in bad shape. She is said to have spent more than ten million pounds restoring it over about a decade.
She also brought books. Over the years she had collected more than 8,000 volumes of English women's writing from 1600 to 1830: novels, sermons, Quaker cookbooks, and the minor and forgotten books that a famous author's work grew out of. She gave the whole collection to the house. In July 2003 it opened as the Chawton House Library and the Centre for the Study of Early English Women's Writing, with a degree-granting partnership with the University of Southampton. Its Novels On Line project made many of the texts free to read online. It is the same move as at Stanford. She took something locked away and connected it to everyone else.
She explained why the collection mattered in terms of context, not genius:
— Sandy Lerner, on the Chawton House collectionAll of a sudden, Jane Austen made a whole lot more sense. It wasn't like she'd dropped in from outer space with six of the best novels in the English language and left. You could see her in the context of things she'd read.
That is not a common view among founders. Silicon Valley usually tells its history as individual brilliance, the genius who arrives from nowhere with an invention. Lerner, who has more reason than most to want the credit, keeps pointing instead to the shared network that individual brilliance comes from. She points to the public protocols, to the shelf of unread women novelists, to Yeager's router and Bechtolsheim's boards. The same instinct that made the love-letter legend untrue made Chawton worth paying for.
The admirers who called her the woman who saved the library also had theories about why Austen. Muscat offered the obvious one: Austen was a creative, intelligent woman stuck in a man's world, and Lerner had been one in the tech industry of the 1980s, which she says was very macho. Lerner has said little to confirm the parallel. She has simply kept reading. In November 2011 she published Second Impressions, a sequel set ten years after Pride and Prejudice, under the pen name Ava Farmer, meaning "a Virginia farmer." Readers who know Austen well said she got the tone right. At the Jane Austen Society of North America's annual meeting in New York in 2012 she gave a talk titled "Money Then and Now: Has Anything Changed?" The room was full, and some of the audience disagreed with her. It is hard to think of anyone better qualified to give it.
In 2015 the Lord Lieutenant of Hampshire, Nigel Atkinson, presented her with an honorary OBE at Chawton House itself. The next year she resigned from the board of trustees and withdrew her foundation's funding. The house, which by then had other supporters, put out an urgent public appeal to stay open. The sources do not explain why she left. The pattern is familiar, though. She builds the thing, connects it, and gives it to an institution. When the institution's direction and hers no longer match, she leaves. She does not appear to stay where she is not running things.
The sequels continued. In 2017 she published Caticons: 4,000 Years of Art Imitating Cats, built largely around her own collection; a cat-news site called it a "Must have." In 2021 she published Archie Goodwin's Office, Nero Wolfe Speaking, a continuation of Rex Stout's detective novels. "I am sure that I will have the Wolfe Pack howling at the door," she wrote. She founded a recording label, Sono Luminus. She compiled The Dilettante's Dictionary of audio-engineering terms. In 2023 she wrote the foreword to a book called Founder vs Investor, about the toxic relationships between venture capitalists and the people they fund. That one was also a sequel, in its way.
The common thread is continuation. She takes characters, buildings, protocols, and breeds of cattle that someone else began, and keeps them going after their creators have gone or been pushed out. Another reading is that she has spent her life doing for other people's work what nobody did for hers at Cisco. She lets it go on, gives it a future, and keeps the original authors' names on it.
Blue Sky and Overhead Wire
In April 2013 CNBC asked her about the technology industry. She said it had stopped thinking about the future. "When I was in graduate school," she said, "all of the big companies had labs, all doing blue sky research." Research without a defined goal is how you get the unexpected discoveries. "Nobody has a long-term view anymore, it's quarter-to-quarter and nobody is willing to fund blue sky research."
She had a warning about timing. "We're still reaping the benefits of the blue sky research from 20 to 30 years ago. In 20 to 30 years, it's not still going to be there." Asked about Google, she was polite and not persuaded: "People are trying to go to the moon; they're trying to do fun things. Blue sky research is about research."
Coming from her, this was a defense of what she had been part of at Stanford: the years when a lab could pull cable without asking permission and give its results away. It was also a criticism of the investor timelines that had pushed her out. Both a quarterly earnings report and a four-year vesting schedule treat time as something to collect on. She treats time as land you protect so that something can grow there later.
Cisco, meanwhile, became exactly the kind of company its founders had not set out to build. It reported $56.65 billion in revenue for fiscal 2025 and had 86,200 employees. It joined the Dow Jones Industrial Average in 2009, replacing General Motors. It spent $28 billion on Splunk. It now sells to governments around the world, and its record includes surveillance lawsuits, NSA backdoor allegations, and layoffs announced in the same week as record profits. The lowercase c is long gone. The logo still shows the bridge.
In 2009 the IEEE Computer Society gave its Computer Entrepreneur Award jointly to Sandra Lerner and Leonard Bosack, "for founding Cisco Systems and pioneering and advancing the commercialization of routing technology and the profound changes this technology enabled in the computer industry." By then they had been divorced for nearly two decades and were still friends and still directors of the same foundation. They accepted together. Every tribute to one of them still mentions the other.
At Ayrshire, among the details Heath reported in 2010 between the price of the turkeys and the size of the cattle losses, there is one that stands apart. The farm did not string overhead wire. Lerner worried that power and utility lines over the fields might cut the wings off the eagles and hawks that dove down to hunt.
In 1982 she pulled coaxial cable across a university campus without permission, to make machines talk that had never been meant to. The sky over her pastures in Virginia had no wire in it at all.
Part IIThe Playbook
The principles below come from the record of a founder who did nearly everything right in building a company and one thing badly wrong in financing it. She then spent three decades and several more ventures trying to keep that mistake from happening again, to herself and to others. Some of the lessons are defensive: governance, legal counsel, the gap between what is good for a company and what is good for the people who started it. Others concern building: how to make money on tools anyone can use, how to turn anger into a product category, how to make a market pay for a mission. Each draws directly on the evidence above.
The principles
- 1.Profitability is leverage. Use it before you sell it.
- 2.The term sheet is the company. Hire your own lawyer.
- 3.You are not the company.
- 4.Assume your investor's incentives can decouple from yours.
- 5.Build the private box on the public protocol.
- 6.Treat operations as part of the invention.
- 7.IP only buys time. Spend it on reputation.
- 8.Turn contempt for an industry into a category.
- 9.Make the market pay for the mission.
- 10.Give it away before it owns you.
- 11.Fund the context, not just the genius.
- 12.When you're in a rut, back out slowly.
⌁
Routes Taken
A timeline of Sandy Lerner's ventures, exits, and continuations.
1975
Graduates from Chico State; finds Claremont's computer lab and soon runs it.
1977
Meets Leonard Bosack sharing time on Stanford minicomputers; earns econometrics M.A.
1982
With Bosack and others, builds an unauthorized Ethernet interface and pulls coaxial cable across campus.
1984
Incorporates cisco Systems on December 10.
1986
Kirk Lougheed joins as employee No. 4; the first dozen routers ship from the Atherton living room.
1987
Sequoia's Don Valentine invests; Stanford licenses router software and boards to Cisco.
1990
IPO on February 16 at $224M market cap; Lerner fired August 28; Bosack resigns.
1992
Foundation backs the restoration of Chawton House in Hampshire.
1996
Buys Ayrshire Farm in Upperville, Virginia.
2000
LVMH acquires Urban Decay (Feb.); Cisco passes $500B in market cap (Mar.).
2003
Chawton House Library opens after a decade of restoration.
2009
IEEE Computer Entrepreneur Award, shared with Bosack.
2011
Publishes Second Impressions as Ava Farmer.
2015
Receives honorary OBE at Chawton House.
2023
Writes the foreword to Founder vs Investor.
Principle 1
Profitability is leverage. Use it before you sell it.
Cisco was profitable from the start, by Lougheed's account, and Lerner says flatly that it "never had a red quarter. Never." It still spent three years looking for venture money. When the money came, Lougheed says, the company "never used" it. Even in its smallest form, eight months and $109,000 in sales, Cisco had what founders almost never have, which was no need for anyone's capital. That made it the strongest negotiating position in the valley. It was given up for "30-odd percent" and a lawyer the founders did not choose.
The mistake was not taking venture capital. It was misjudging what they were buying. Lougheed's honest answer is that Sequoia supplied "people who knew how companies work," meaning governance experience, recruiting, and credibility. Those things have real value, and they can be priced. A founder who does not need the cash should negotiate as though the investor is buying in, not rescuing the company, because that is the actual situation.
"If you can," Lerner told founders in 2012, "fund your venture by yourself." Read that less as a refusal of outside money and more as a statement about timing. Self-funding is valuable mainly because it lets you choose the moment, and the terms, on which you stop.
Tactic: Before your first priced round, write down what the investor provides besides cash, put a number on it, and negotiate as a profitable company that is selling access, not as a startup asking for help.
Principle 2
The term sheet is the company. Hire your own lawyer.
"We used Don's lawyer." Lerner has repeated those four words more than almost anything else she has said in public, and with good reason. The four-year vesting schedule, the missing employment contract, and the control that came with the stake all came out of documents drafted by the other side's counsel. "Don't buy lines like, 'You guys are busy; we'll just have someone draw up some papers, and it will be very pro forma,'" she warned. "Yeah, right."
The Cisco founders were not careless people. They had built a router that worked across incompatible protocols. They simply did not see legal architecture as architecture. A shareholder agreement is a routing table. It decides where control goes when conditions change, and somebody else wrote theirs.
The cheapest point at which to fix this was before signing. Separate counsel would have cost a small fraction of the founders' eventual $170 million, and it might have produced the one clause they did not get: protection against being fired.
Tactic: Retain counsel who works only for you before any term sheet arrives, and treat "it's pro forma" as a sign to read more closely.
Principle 3
You are not the company.
"Len and I believed a company and its founder are the same. They absolutely are not." This is Lerner's most painful lesson, and probably her most widely applicable. She made "every decision based on what was good for the company," and by her own account it "pretty much ruined my marriage and my health." Then the company, acting through its board, decided it was better off without her.
The belief that founder and company are one feels like loyalty. In practice it leaves you exposed. A founder who has merged her identity with the company has nothing left to bargain with when the company's interests change. The board can see the difference between them perfectly well. Only the founder cannot.
The comparison with her later ventures is telling. She invested in Urban Decay through a venture vehicle, & Capital Partners. She gave Chawton House to an institution and left its board when she disagreed with it. She sold Ayrshire when the "house phase" was over. After Cisco she never again treated a venture as though it were her body.
Tactic: Write a short memo to yourself every year listing where your personal interests and your company's interests have diverged, and act on what it says.
Principle 4
Assume your investor's incentives can decouple from yours.
Lerner learned her model of investment on a cattle ranch. If you put money into the animal, you want the animal to do well. "My family had a small business. I always thought that if someone invested in your business, that meant he or she believed in it." Her mistake was not naïveté in a general sense. She was using a model that holds when owner and operator are one person, and venture capital separates the two.
"I did not realize he had decoupled the success of the company from that of the founders." By the measure investors care about, Cisco without its founders did very well, reaching half a trillion dollars ten years later. That outcome does not show bad faith. It shows that an investor's goal of returns and a founder's goal of building "the best networking computer that you could build" can point in different directions while both parties act in good faith. "They're not the same," she said.
Her conclusion in 2013 was careful: "I don't believe all VCs are adversarial." She was not saying investors are villains. She was saying a founder should plan for the scenario in which keeping her becomes optional.
Tactic: Map in writing the conditions under which your investors would gain from removing you, and negotiate protections for those specific scenarios before you need them.
Principle 5
Build the private box on the public protocol.
"In Cisco, we made a lot of money on public protocol." The protocols that made multiprotocol routing work had been developed collaboratively and released into the public domain. Cisco did not own the language networks used. It owned a dependable, shippable machine that spoke all of them.
This is a more general strategy than it looks. Any open standard creates demand for whoever makes it usable. The money lies in the gap between what a standard allows and what a customer can actually deploy on a Tuesday morning. Cisco's boxes went out by UPS from a living room to universities and HP because they worked across environments no one had designed to work together.
Lerner later applied the same reasoning to social media and to Amazon. She called these channels, "like radio or TV," and said the winners would need "a significant technical component of intellectual property" on top of the shared layer. Being built on open tools is not a weakness. Having nothing proprietary on top of them is.
Tactic: Find the open standard your customers already depend on, and build the specific, supported, reliable implementation they cannot easily build for themselves.
Principle 6
Treat operations as part of the invention.
Bosack built the routers. Lerner, by the IEEE's account, "headed up all departments except engineering," including the customer-facing operation she later ran as vice president. Routing technology existed at Stanford before Cisco did. What did not exist was a company that would stand behind a router the way a utility stands behind the electricity it supplies.
This work is mostly invisible, and it is often why a technology spreads. Lab equipment becomes infrastructure when it ships reliably, when someone answers the phone, and when the documentation makes sense to an engineer at Ohio State. The credit disputes that followed Cisco, from Yeager's complaints to the Mercury News article in 2001, were always about who invented the machine. Nobody fought over who made it something a business could buy.
Founders with technical backgrounds often undervalue the people who do this work, and sometimes they undervalue it in themselves. It is a different kind of engineering, applied to trust instead of to packets.
Tactic: Assign one founder or senior leader the explicit job of making the product deployable and supportable at scale, and give that work the same status as building the core technology.
Principle 7
IP only buys time. Spend it on reputation.
"Even if you have that technical IP, it will only buy you time before someone reverse engineers it. But what you can do is buy yourself time." Lerner treats intellectual property as a lease on the future that will run out. The real question is what you build before it does.
Urban Decay had no intellectual property worth mentioning. Cosmetics, in her description, is "a relatively small kit of tools and everyone is playing with the same one." The company sold because, over six years, it built "brand cred," a reputation for being "really good makeup" behind slogans that sounded like a provocation. LVMH paid for that reputation. A competitor could have copied the color names. It could not have copied six years of trust from customers.
Seen this way, IP and brand are two forms of the same asset, time already invested. Use the first to get the second.
Tactic: Estimate how long your technical advantage will last, and set a deadline within that window by which customers must trust your name more than your feature list.
Principle 8
Turn contempt for an industry into a category.
Urban Decay began with an aunt's suggestion and a bad mood. "I was just pissed off that cosmetic companies were telling women they had to look like Barbie." Lerner did not take the anger to a focus group. She put it straight into the product. She named the colors Smog, Mildew, Roach, and Gangrene, and wrote the tagline as a challenge: Does pink make you puke?
The approach works because strong dislike of an industry's assumptions tells you where demand is going unmet. If you find the pink wall insulting, other customers probably do too, and no incumbent will serve them, because serving them means admitting the pink wall was a mistake. The Mattel lawsuit threat over Burn, Barbie, Burn was a sign the company had hit something real.
Cisco had the same origin in a gentler form. A university that couldn't approve a commercial router in "less than years" frustrated two people into starting a company to sell it.
Tactic: Write down the industry assumption that most irritates you as a customer, then design the first product so that it openly contradicts that assumption.
Principle 9
Make the market pay for the mission.
At Ayrshire, Lerner's theory was that heritage breeds would survive if there was a market for them. If people would pay for Highland cattle and heritage turkeys, the market would keep those bloodlines alive in a way donations could not. So she ran it as a vertically integrated business: farm, store, tavern, catering, carriages, and wholesale accounts with restaurants like the Inn at Little Washington.
She was honest about the numbers. A $230 turkey still lost money. The operation lost about $1 million in 2010, mostly on cattle. "I can sustain the loss indefinitely. But I don't want to." That distinction is the principle. A mission paid for out of a fortune is a hobby. A mission the market pays for can last beyond the person who started it.
The certifications, first in Virginia for Certified Humane and Certified Organic and later Predator-Friendly, served this goal too. They turned values into signals a buyer could rely on, which is how a mission becomes a price premium.
◆
Founder Wisdom vs. Lerner's Record
Where conventional startup advice meets what actually happened.
| Conventional wisdom | Lerner's evidence | Signal |
|---|---|---|
| Raise venture capital to grow. | Cisco was profitable before Sequoia and "never used" the money. | Reprice |
| Standard docs are standard. | "We used Don's lawyer" — and got no employment contract. | Risk |
| Founders embody the company. | "They absolutely are not." Fired 193 days after the IPO. | Risk |
| Defend your IP at all costs. | IP "will only buy you time"; Urban Decay sold on reputation. | Reframe |
| Missions need donors. | Heritage breeds survive if a market pays for them. | Scale |
| Hold your equity for the upside. | Sold everything and gave 70% away; Cisco hit $500B anyway. | Tradeoff |
Tactic: Separate your mission's costs into what customers will pay for and what you will subsidize, and then work to move items from the second list to the first.
Principle 10
Give it away before it owns you.
The couple gave about 70 percent of their Cisco proceeds to charitable vehicles almost immediately, including a reported $34 million endowment for the Bosack/Kruger Foundation. It was an early and deliberate decision. "I made a very conscious decision not to let it ruin my life. I think money is a good tool and I also think it is a drug."
The strategic point is that money left unassigned accumulates expectations, both other people's and your own. By committing most of it early to defined causes (animal welfare, conservation science, early women's writing), Lerner kept the remainder small enough to be useful. It funded Urban Decay, Ayrshire, and Chawton, ventures chosen out of interest and not for status. Her description of the rest, "in a big pile" that someone else sweeps up, is funny and also a real stance. She does not let the money set her agenda.
The cost was real. Cisco's market value at its peak makes the decision to sell everything look painful in hindsight. She seems to have decided that the alternative, staying tied to a company that had fired her, cost more.
Tactic: Decide in writing, before a liquidity event, what share of the proceeds will go to defined purposes, so the decision is made before the money starts making it for you.
Principle 11
Fund the context, not just the genius.
Lerner's justification for Chawton is her clearest statement about where ideas come from: Austen "wasn't like she'd dropped in from outer space," and "you could see her in the context of things she'd read." The 8,000 volumes she collected were mostly not masterpieces. They were the surrounding material that made the masterpieces understandable.
The same idea runs through her criticism of the technology industry. "We're still reaping the benefits of the blue sky research from 20 to 30 years ago," she said in 2013, and she warned that the harvest would run out. Cisco itself came out of that environment: public protocols, Yeager's router, Bechtolsheim's boards, ARPANET machines, and a university network many people had built. Breakthroughs need that surrounding work, and someone has to pay for it before anyone knows what it will produce.
For operators, the lesson is to set aside a budget, however small, for exploration with no specific deliverable, and to keep a record of the work that made your own success possible.
Tactic: Set aside a fixed share of your R&D or learning budget for work with no defined deliverable, and protect it from quarterly review.
Principle 12
When you're in a rut, back out slowly.
The field at Ayrshire is a small parable, and Lerner's career is the larger one. After Cisco, she said, "I spent years crawling out from that." She did not try to start another Cisco or win the old fight. She went into cosmetics, then cattle, then Austen, then Nero Wolfe. Each change was a reversal out of the previous rut onto a path she had chosen herself.
Her explanation of her own story is useful here. Everyone assumed the hard move was from Silicon Valley to farming. She insisted the hard move was the earlier one, from the farm to Cisco, and that going back was "the easy part." Getting out of a rut often means retracing your steps to the place where you last had traction, not finding a new direction.
Reinvention is often imagined as dramatic. Hers looked like careful driving: shift gears, keep the wheels turning, back up an inch at a time, and keep going once the tires grip.
Tactic: When a venture stalls, identify the last point at which you had real traction and work back toward it step by step before trying anything new.
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
— Sandy Lerner, 'Investors Aren't Friends' (2013)When I was fired, it was devastating. I spent years crawling out from that. I did not understand an investor could be an adversary.
— Sandy Lerner, to ZDNet, on the Cisco yearsThe venture people were into building a bigger business and we were into building the best networking computer that you could build. They're not the same.
— Sandy Lerner, to ZDNetI have yet to meet a rich and powerful animal. You know, they don't vote and they don't talk. They're such a part of our life. We use them. We eat them. We make money from them. We exploit them. Somebody needs to watch out for them.
— Sandy Lerner, in Something Ventured (2011), asked if she is eccentricI am now that I'm rich. I used to just be weird.
— Sandy Lerner, to CNBC (April 17, 2013)Nobody has a long-term view anymore, it's quarter-to-quarter and nobody is willing to fund blue sky research.
Maxims
- Profit is a bargaining position. A company that doesn't need your money should never negotiate as though it does.
- Get your own lawyer. The person who drafts the documents decides where control goes when things change.
- You are not the company. Founders who merge their identity with the firm have nothing left to negotiate with when the board decides otherwise.
- Investment can decouple. An investor can want the company to thrive and the founder to leave, in good faith.
- Sell the box, not the protocol. Open standards create demand for whoever makes them reliable to deploy.
- IP only buys time. Use the lead to build a reputation competitors can't copy.
- Anger is market research. The industry assumption you most resent is often a customer segment no one serves.
- Make the market pay for the mission. Things a subsidy keeps alive will die with the subsidy; things a market pays for can last.
- Money is a tool and a drug. Decide where it goes before it starts deciding for you.
- Back out of ruts slowly. Reinvention usually means returning to the last place you had traction.
