The Disruptor In The Valley
Justin Kan and Emmett Shear watched their first startup, an online calendar called Kiko, implode when Google decided to do the same thing in 2006. They sold Kiko's scraps on eBay for $258,000 and wondered what to do with their lives. So the pair did the only thing they could think of: They went to see Paul Graham at his house in Cambridge, Mass., near Harvard Square. Graham sat them down and helped bang out a plan to create Justin.tv, now the Web's biggest portal for live video, with 31 million users a month and staked with $7.2 million of venture capital.
Justin.tv is hardly the first, nor the last, company Graham has sent sprinting. Graham is the father of Y Combinator, a startup-rearing juggernaut that's part incubator, part drill sergeant and part liaison to the investor class. Y Combinator--a computer term for a program that runs other programs--has fired up 200 companies since 2005, jarred the balance of power between entrepreneurs and Silicon Valley's elite money, and chiseled a new paradigm for launching technology companies. Graham's formula: Get up and running (bugs and all), gather feedback, tweak and grow.
YC's three-month boot camp for startups, run twice a year in Mountain View, Calif., attracts 1,000 applicants for roughly 40 spots. Graduates are expected to emerge with a working product, customers and revenue. They also get a crack at pitching their ideas to investors on Demo Day, an event that lures venture capital's Sand Hill Road crowd and every prominent angel investor in the Valley.
YC puts up $11,000, plus $3,000 per founder, for each company in return for a piece of pure equity of around 5%. That equity could be worth real money should the companies take off. A high price for founders, perhaps, until you see scores of venture capitalists and angel investors jousting to pay handsome premiums for companies bearing the YC stamp. Of the 36 startups in YC's recent class, ended in August, 30 have raised fresh capital, many of them over $1 million.
"We didn't mean to invent this new model," says Graham, who at 45 has sandy hair and a youthful earnestness. "It all happened by accident."
The accident was a summer program Graham started in 2005 for college students who were tinkering with business ideas. Instead of working a boring internship at a big company, Graham's pitch went, win $5,000 to work on your startup in Cambridge with guidance from Graham and his friend, MIT professor Robert Morris--two guys who launched Viaweb, a maker of software that built storefronts online, and sold it to Yahoo for $50 million in 1998.
"It was supposed to be a throwaway project for these students," recalls Graham. "By the end of the summer we were like,"'Whoa, we've got something here!'"
Graham received 400 applications for the summer program. Of the 8 he accepted, 4 had blossomed into serious ventures by the end of the summer: Loopt, a social-mapping ser vice, now with 4 million users; Reddit, a user-aggregated news site acquired by Condé Nast in 2006; TextPayMe, a mobile payment service bought by Amazon in 2006; and Kiko, thwarted by Google.
Y Combinator's influence in Silicon Valley has burgeoned ever since. Some refer to its growing network of graduates as the YC mafia. They protect their own, collaborate and, to a person, regard Graham as their sensei. Some go on to be investors and mentors in their own right.
"The right advice has always been more important than money," says Greg McAdoo, a partner at Sequoia Capital, which has invested in Google, Yahoo, PayPal and YouTube. "But nobody has been able to do it on this kind of scale before."
Jealous types claim Graham runs a fund to pick off YC's best as soon as they leave the womb. YC has raised two pools of funding from outsiders--$2 million in 2009 and $8 million in 2010, from the likes of Sequoia Capital and prominent angel investors Paul Buchheit and Aydin Senkut--but that money is only for YC's small initial equity investments. Graham concedes that individual YC partners have invested in a few startups that hadn't been able to attract much outside funding. But, he insists, YC startups have no obligation to accept additional capital from anybody.
Graham grew up outside of Pittsburgh, where his father, a physicist, designed nuclear reactors and his mother raised Graham and his sister. He started writing computer code in high school; one program predicted the flight paths of model rockets. Graham would eventually earn a Ph.D. in computer science from Harvard, where he intended to concentrate on artificial intelligence.
Disenchanted by the prospects of ever building a truly intelligent machine, Graham moved on to painting, attending the Accademia delle Belle Arti in Florence and the Rhode Island School of Design. He jammed into a tiny New York apartment to start his art career and was often broke. "I decided to go out and solve my money problem for good," he recalls.
Viaweb did that. Free and flush after the sale, Graham found a creative outlet writing essays and posting them to his Web page. He opined on esoteric programming issues and more accessible topics, such as "Why Nerds Are Un popular." One of his essays, "Beating the Averages," which praised Lisp, a programming language that helped Graham build Viaweb, snared 50,000 page views. "All of the sudden, I was writing for a lot of people, and that made me want to write more," he remembers.
Another essay, "How to Start a Startup," based on a talk he gave at Harvard, got him thinking seriously about angel investing and eventually inspired YC's formation.
Graham's three fellow founders--Morris, Jessica Livingston and Trevor Blackwell--were close confidants. Graham met Morris, an authority on the Unix computer language, at Harvard. When Bell Labs, where Unix was developed, wanted to integrate Web programming standards into Unix, it called Morris, then age 17. Livingston wrote the book Founders at Work, a collection of profiles of marquee technology entrepreneurs. (Graham was dating Livingston when YC began; they're now married.) Blackwell, another Harvard Ph.D., worked at Viaweb and later launched Anybots, which made the first walking robot that dynamically adjusts its weight distribution, like a human, as it moves across changing surfaces. The partners recently added Harjeet Taggar, who sold his startup after being part of YC's winter 2007 class.
Graham reckons the next crop of applications, due in late October, could hit 1,500. About 80 will earn the right to a nerve-racking interview with the founders in Mountain View. Many of the best applications come from accomplished programmers who have working product demos but lack gilt-edged Valley Rolodexes. But Graham and company aren't just looking for nifty concepts. "We realized early on that the founders matter more than the idea," Graham says.
YC's vetting process plays a big role in attracting investors. "You're getting prescreened deal flow," says Ron Conway, a prominent Valley angel who invested in Google, PayPal and Twitter. Conway has put capital into 20 YC companies, including 7 from the latest class. Investors are also enticed by the $200 billion of cash piled up on the balance sheets of Apple, Microsoft, Google, Cisco, Oracle, Intel and Qualcomm. With any luck, some YC grads may soon get snapped up.
For startups the YC experience revolves around Thursdays. During the afternoon entrepreneurs meet for scheduled office hours with Graham or one of the other partners. They report their progress and brainstorm how to solve nagging problems. Scribbling bullet points on a white board, Graham recently implored Francis Duong and Edmond Yue of TapZilla, which offers one discounted iPhone app at a time, to get a new deal up daily. The pair had, at that point, put up 10 apps in the span of a month, selling 10,000 in all. Graham wants TapZilla to become a site where app hounds feel compelled to check in every day.
YC recruits say Graham has a knack for cutting straight to an idea's weakness or for amplifying its unique strengths. "His brain is a giant warehouse of startup failures and successes," says Rudy Adler, cofounder of 1000Memories, a website where people can celebrate deceased friends and family members.
Adler's first idea involved a city-mapping site that tagged friends' favorite places. Graham was leery, having seen failed startups go down a similar path. Still, he let Adler and his mates give it a shot for a few weeks before refocusing them on 1000Memories, their auxiliary plan.
Startups who continually lose their way can lose Graham, too. "With Paul, you're either in or you're out," says one YC graduate. "You don't want to be out."
A home-cooked meal follows office hours. A recent dinner featured 60 pounds of chili, made in seven Crock-Pots, served over rice. Cookies followed. A guest speaker--often plucked from technology's A-list, such as Facebook's Mark Zuckerberg and Groupon founder Andrew Mason--spins tales as people eat.
The avalanching interest in YC is on full display during its Demo Day. YC's warehouse space, corrugated with three rows of tables made from plywood and white melamine, overflows with angels and venture capitalists. (The space can accommodate only 150 investors at a time, so there are now three Demo Days per class.) Each startup gets two and a half minutes to make their case, show their product and leave a good impression.
The event is invitation only, but Graham says he allows in any investor who asks. Demi Moore and Ashton Kutcher, who has invested in several YC companies, showed up for the most recent gathering in August. Kutcher is just one of a growing group of 32,000 U.S. angel investors (so-called accredited individuals with more than $1 million in assets) who last year wrote checks totaling $12.4 billion, says Scott Shane, an economics professor at Case Western Reserve University.
"It's all about hype," says Laurence Albukerk, managing director of EB Financial Group, a Valley firm that facilitates trading stock in young companies. "The more hype you get, the higher your valuation."
Plenty of Valley insiders could do without all that hype, especially when the result is having to pay more for stakes in YC's offspring. "You show up and a lot of other VCs are there and everybody has good pitches, so things can get bid up," says Richard Heitzman, managing partner at FirstMark Capital, which put capital into StubHub, Netgear and TheStreet.com. "But in the end people are going to pay what they're going to pay. I don't begrudge them that value creation."
The real winners are rising stars like Jessica Mah. "The competition this created for us was amazing," says Mah, a 20-year-old summer YC grad and founder of Indinero, a cloud-based accounting application for small businesses. (Mah graduated from Berkeley with a degree in computer science at age 19.) Indinero just raised $1.2 million in fresh capital at an undisclosed valuation. Mah says it was three times as much as Graham figured it would be.
Indinero is Graham's archetype. Mah had a working prototype when she applied to join YC for the summer. Graham pushed her to add features, polish the site and incorporate the changes quickly. "Put it out there and let users decide," he says. At the end of August Mah had 2,500 customers; two months later she was up to 6,000. Critics piped up quickly, convincing Mah to make the interface more intuitive. "Watching somebody's blood pressure rise as they try to use your product can be enlightening," she says.
Being able to prove a business concept in a hurry, even if the product isn't perfect, is especially attractive to investors, says Peter Bell, general partner with Highland Capital Partners, a venture firm in Lexington, Mass. that staked Digg, Mapquest and Ask Jeeves. "The power has always been with the entrepreneur if they have a good team and a good product," he says.
Other investors aren't so sanguine. Michael Arrington, editor of the influential blog TechCrunch, tipped the pot on a simmering controversy in September when he accused a prominent group of angel investors (whom he calls "super angels") of colluding to keep competitors out of deals and hold valuations down, partly in response to Graham's newly empowered students. Arrington crashed a dinner at Bin 38, a haute San Francisco eatery where he says the group of investors was hatching its plan. "Mike is wrong--he wasn't there," says Dave McClure, an established Valley investor and one of the angels at the meeting. (Others who were present deny the allegations as well.) McClure does acknowledge YC's effect, however. "Paul has done a great job coaching his guys on the investment side of things, and there is a lot more competition for these deals now," he adds.
Graham remains suspicious of the group but finds the whole dustup beside the point, given the overall size of the investor pool. "They couldn't have ever achieved anything, so the whole thing is really kind of comical," he says. Arrington disagrees: "Deals rarely get done without a few of these guys being involved," he says. "In layman's terms, what they were doing can definitely be considered collusion."
For his part, Graham is always pushing better financing terms for entrepreneurs. One useful tool: convertible notes that turn into equity upon a startup's next valuation. The notes often come with valuation caps to protect early investors from being diluted should a company hit it big.
Say a convertible note worth $1 million was equal to 50% of a startup's estimated worth at the time the note was written. Now say the company takes off and the next time it raises money it is valued at $100 million. The $1 million of debt becomes $1 million of equity, but that stake is worth just 1% of the company. However, if the note came with a valuation cap of $10 million, that means it would be worth 10%--or $1 million divided by $10 million--of that $100 million company. The caps give entrepreneurs more flexibility, allowing them to pit investors against each other by offering early backers lower valuation caps. "Convertible notes mean investors can no longer drag their heels," says Graham.
Like any good business, Graham's is drawing competition. David Cohen, chief executive of TechStars, an incubator based in Boulder, Colo. with satellites in Boston, Seattle and New York, says 100 similar outfits, mostly small, now exist across the U.S. TechStars accepts ten startups per year in each of its four markets. Those that make the cut stay for three months; 60% of its first 40 graduates have each raised more than $600,000 after leaving the program.
After Cohen's group there's a clear dropoff. But more big players are ramping up their own versions of YC. Chicagoan Eric Lefkofsky, the cofounder of Groupon who's flirting with billionaire status, is planning something in his city soon.
Graham welcomes the competition. "It's better for entrepreneurs," he says. "If we make a mistake and don't fund somebody great, they still have a chance."