Among the Summer Founders

In the summer of 2005, I pitched a story to Inc. magazine about a quixotic experiment taking place in Cambridge, MA, being run by an entrepreneur named Paul Graham. The magazine agreed to send me to cover the first-ever Y Combinator class—then never got around to running the story. It’s fun to read now as a bit of a time capsule: How many people remember that Reddit was founded that first summer, or that Y Combinator’s current CEO, Sam Altman, was a class member, as was the late Aaron Swartz?

Summer Founders

Once upon a time, young men and women used to pay to go to something called "business school." If they were wealthy, their parents paid a ransom called "tuition," and if they were not, they took out personal loans running to hundreds of thousands of dollars. But the strangest part is that back then not one business student ever turned a profit. Hard to believe, but that's the way it was, even as recently as the turn of the 21st Century....

Paul Graham didn't set out to reinvent the system by which businesses are made and business people are educated. Graham, a painter and philosophy major in college, started ViaWeb with his partner Robert Morris out of a fleabag apartment in Cambridge where the cracks in the walls were plugged with tinfoil. They sold the company to Yahoo! in 1998 for $49 million.

Like a lot of tech millionaires, Graham then put his personal wealth to work as an investor. He also bought himself a little cinderblock building on a leafy street in Cambridge, to use as an office and painting studio, walking distance from Harvard, where Morris was now an Associate Professor. Graham became something of a gentleman dot-commer, writing and lecturing widely on software and startups, publishing smart, literate essays with irreverent titles like "Why Nerds Are Unpopular" and "A Unified Theory of VC Suckage" (Venture Capitalists can take a drubbing in Graham's writings). His essay collection, Hackers & Painters, is a definitive guide to the nuanced culture of the tech world, covering everything from why nerds are unpopular in high school to why Lisp is a better programming language than C++. If you're a computer science major at MIT or Stanford, Graham is your J.D. Salinger.

Which is why, when Graham showed up at the Harvard Computer Society to give a lecture one evening in May 2005, the audience was packed with young, aspiring software gurus.

"Hiring is Obsolete!" Graham told the assemblage of nerds-in-training.

For the next hour, he exhorted them to eschew the temptations of the new crop of cushy programming jobs at Google and Yahoo!, and called on them to take a chance on their own terms-to get out there and found startups. And should they fail, he noted, as nine out of ten startups eventually do, grad school will still be there.

To a group of kids who had cut their teeth on adventure stories from the wild frontier but were now graduating to find themselves facing a far blander and more "mature" software industry, and a tighter capital market, this invitation to sign on to a new Declaration of Independence was heady stuff.

Several groups of students approached Graham excitedly after the lecture to pitch their business ideas and ask him for concrete advice on how proceed next. Listening to them, Graham found their ideas so promising that he was tempted to fund them all himself, there and then.

As it is, he actually waited several days.

*

Chris Slowe and Zak Stone were among the Harvard grad students who attended Graham's lecture. Eight weeks later, they were taking a break from intensive product development at their new software company to share Tuesday dinner with Graham and members of the seven other teams to receive funding as a result of that single "Aha!" moment.

On Tuesday afternoon, Graham spent an hour filling two industrial-sized crockpots with the makings of his signature "Glop," a healthy concoction of beans, veggies and some kind of stew meat, served over rice and eaten at a string of plywood tables long enough to accommodate an Amish barn-raising crew. The kitchen and dining room were part of the office/loft that used to be Graham's painting studio and now served as headquarters for the Summer Founders Program.

As members of the program, Chris and Zak got $6000 each, help with nuts-and-bolts like incorporating in Delaware, lots of free advice and all the Tuesday glop they can eat. Their six week-old company, Memmap, Inc., is designing new file management software for PCs-something to replace the look and feel of the average computer desktop.

"We're going to break out of the icon grid," Chris said. "Enough folders! That metaphor is thirty years old now."

The two had started noodling on an idea for file management software a year ago, but hadn't thought of what they were doing as a product, let alone the basis for a startup, until meeting Graham. "I just wanted to design something to make it easier to work on my thesis," Chris explained. "And I'll still be happy if that's what I end up with."

Which is exactly why Graham thinks college kids are so well positioned to start companies: they're highly risk tolerant. Without spouses, kids, mortgages or car payments, students can afford the highly educational fool's-errand that the startup experience can sometimes prove to be. Worst case, it's a paid summer internship; best case, you're the next Google. And everything in between is gravy.

From his own experience, Graham knew that getting funding at the seed level was haphazard even for pros. "When Robert and I started ViaWeb-well, if anyone ought to have been well connected it would have been us--but both of our sources of seed money ended up being people I knew from the art world rather than the computer world," he said. "A guy a friend had moved some art for was the first to come on board. The other guy was a client of my art teacher's husband. If you're lucky you'll get your seed money from Steve Wozniak, but you're more likely to get it from your friend's rich uncle who made his money from starting a shoe store."

And venture capital? Forget about it. Then, as now, venture capitalists wouldn't look at projects seeking less than three or four million dollars--way more than your average software startup needs or can even handle. The randomness and inefficiency of this informal funding system had long galled Graham, but it also suggested an under-served capital market-and the Summer Founders Program is a first attempt at formally exploiting that opportunity.

After that evening at Harvard, Graham called on his old ViaWeb partner, Robert Morris, and they quickly hatched an idea: they would set up a competitive application process and select six to ten teams to fund for a summer of intensive startup activity, based in and around Graham's office in Cambridge. Graham and Morris enlisted two more friends, Trevor Blackwell, founder of Anybots, and Jessica Livingston, a marketing executive from the finance industry, and the four partners quickly formed Y Combinator, an investment fund focused specifically on providing the earliest stage of seed capital to the youngest entrepreneurs. The name comes from a term used in computer science--a Y Combinator is a program that runs other programs. This would be a startup company in the business of starting companies.

Before the ink was dry on the partnership they published an announcement on heavily trafficked industry sites like Slashdot.com, that read, in part:

We give you enough money to live on for a summer, as with a regular summer job. You get to work on real problems, as you would in a good summer job. But instead of working for an existing company, you'll be working for your own; instead showing up at some office building at 9 AM, you can work when and where you like; and instead of salary, the money you get will be seed funding.

With an application deadline only ten days off, it seemed like a shot in the dark, but over the next week 227 applications flooded in. The partners spent an intensive weekend pouring over the entries, from which they chose 21 viable candidates, then flew the teams to Cambridge the following weekend for personal interviews. The process was intense and intuitive.

"We interviewed them on Saturday and Sunday," Livingston said. "And on Sunday evening we made our decisions. It was a verrry long weekend."

The applications that came in told the partners a lot about why young people don't always get funded. "One thing that surprised us was that very smart people often didn't have very good ideas," Graham said. "There was a lot of duplication in the applications. What we got was a huge number of ideas for social networking sites. They came up those ideas because that's the software they use. It's like the freshman autobiographical novel--the social networking idea." Graham is an enthusiast and an optimist, but he's no romantic. "A good idea for a startup," he says, "is something someone will pay a lot of money for."

So Y Combinator decided early in the process that they would be different from most Venture Capital funds in one more respect: they would be funding people, not ideas. Steve, a University of Virginia grad, was one of those who pitched an idea for social networking software, and the partners weren't the least bit impressed with the product--but they were with Steve, and offered his team the chance to come to Cambridge and work on something else-maybe a newsreader? Steve's company, Reddit, was beta-testing its news aggregator by late July, and he was just as happy to be working on anything at all with Graham, who is one of his heroes.

The concept was very simple: In exchange for investment capital and help with the costs and hassle of incorporation and other legal matters, Y Combinator received common stock in the newly-minted company based on an agreed-upon valuation. "Our subscription agreement is 10 pages long, not 50," Graham said. "We don't negotiate terms, because we can't afford the legal fees. You can't scale an elephant down to the size of a mouse without changing the design of the animal. 'Cheap, fast and out of control'--sometimes I feel like that ought to be our motto."

Everybody got the same offer--and the same stipend--$6000 per kid. This really was like a paid summer internship, save for the remote possibility that everyone involved could get filthy rich someday.

Less than a month after the Harvard lecture, Y Combinator had offered funding deals to eight startup companies--not all of which even had a product--the stipulation being that the founders had to come shack up in Cambridge for the summer. How young can you go? Graham had wondered. He was about to find out.

At 26 and 28 and 23, Chris and Zak were among the oldest founders in the new program. The youngest, Aaron Swartz, was18, a rising sophomore at Stanford. Aaron also had the highest profile coming in, being well known in hacker circles for co-authoring an important piece of code used to syndicate blogs and newsfeeds-at the age of 14. He didn't even have a team assembled when he applied, and found his business partner, Simon, online. A student in Denmark, Simon flew to Cambridge from Copenhagen to join Aaron. The two shared a dorm room at MIT, where they worked on Infogami, their next-generation blogging software day and night and subsisted on a diet of dry Cheerios (Aaron explained to me the science behind this regime, saying he had identified himself as a “supertaster,” someone with taste buds too sensitive to handle all but the blandest foods).

"He writes like he's 30," Graham said, describing Swartz. "Yet one of the things they have to do this summer is come here to Cambridge, which means they have to rent an apartment. He's never rented an apartment! Isn't that wild?"

In week five, the teams gave their mid-program product demos and then everyone voted on which companies they'd personally invest in. Another of the youngest founders, Sam Altman, emerged as the clear winner and was subsequently dubbed "most likely to be the next Bill Gates."

Jessica Livingston (whose fearlessness in face of red tape earned her the nickname "The Incoporator" among the kids) said that Sam had been having a bad week.

"Sam was sitting around looking very down," Livingston said. "And when we asked him why, he said it was because today was his 20th birthday. He was all broken up over turning 20, and how he felt he hadn't accomplished enough, fast enough."

A week later, Altman flew to Chicago, where he signed Sprint on as the first client for his company, which uses the GPS chip built into cell phones to help groups of people (like college friends, or members of a team) find each other using mapping software.

"That's what has really struck me about this summer," Livingston said. "Here you've got Sam Altman, who has the skill and the confidence to fly to Chicago and ink a deal with a major corporation, but when he gets there, he isn't old enough to rent a car to get to the meeting."

At the other end of the spectrum from Aaron and Sam were the "Mikhails," a group of three Russian-born students who cultivate an aura of near-comical nonchalance (two are actual Mikhails, while the third is a Greg). They heard about the program the Friday before the Saturday deadline, and fired off an idea. "We have started companies before," shrugged Mikhail #1. "We had lots of ideas lying around."

The one they chose was a subscription-based application to help online advertisers monitor click fraud. "We picked one we could do quickly and well," said Greg. And apparently they chose well. Their product is finished. "We're on to marketing now."

"They've already gotten their company written up in the Wall Street Journal," Livingston said in an aside, shaking her head. "We're not even sure how they did that. They're mysterious."

If one team stands out as typical among this group of eccentric characters, it would have to be Emmett and Justin, the founders of Kiko, who best fit the mold of young aspiring internet mavericks. If this were a John Hughes teen comedy (which it could easily be, what with the Russians are mysterious, a Cheerio-eating prodigy and a handsome team working on a dating site), Emmett and Justin would be the main players, the ones the audience identifies with. Best friends since they were two years old growing up together in Seattle, they started working on Kiko last year at Yale.

"We were sitting in a bar," said Justin. "It was senior year. We realized we had the best access to intellectual capital we'd every have in our lives, and we'd better do something."

Justin had a programming job lined up for the fall, but said he now knew he wasn't going to take it, whatever happened with Kiko. "I can't imagine anything better than working for yourself," he said.

"If worst came to worst," Emmet said, "We'd go back to Seattle, live with the folks, work out of the garage."

This is what Graham means when he declares that "hiring is obsolete." The most important thing Y Combinator is inculcating in the Summer Founders the sense that they can indeed bootstrap their way into the market. Six weeks earlier, if you'd asked any of the Founders what they needed most to start a software company, the answer would have been capital investment. But by the time they welcomed the Tuesday night guest speaker, Venture Capitalist Stan Reiss of Boston-based Matrix Partners, they had all been thoroughly schooled on Graham's "Universal Theory of VC Suckage."

Graham doesn't hate venture capitalists, not by a stretch (Reiss came because he and Graham are friendly), but he does think that they're largely irrelevant to fostering innovation in the software industry, in part because they have too much money. "There's a reason existing VC funds don't want to get into this end of the business," he says. "They manage hundreds of millions of dollars, yet they only have so many partners." Each partner can only oversee so many startups, so that the minimum it makes sense for them to invest is $3 to $5 million.

Graham believes that $3 million is a soul-killing amount of money for your average software startup. "Getting the $3 million hoses a lot of companies. Why? Because you have to spend it! Suddenly you're hiring lots of people, when maybe what you really need is for your group of founders to keep plugging away. That kind of money, for a small startup, it's like taking steroids, it makes you bigger but it can also kill you."

After dinner, Reiss sketched out the realities of VC funding, explaining what kind of companies investors look for, what the pros and cons of venture funding are for a startup, how to design an exit strategy. When he finished, they jumped in eagerly:

"What size should you be before you look for VC money?"

"You said VCs offer more help to a company than angels. What does that mean?"

"When's the last time you actually funded a company run by three guys just out of college?"

"How often is it the first people with an idea who get the hit, and how often is it the next guy who succeeds?"

"How much revenue do you expect a company to be generating when you invest?"

"What can a small company do to protect itself from being crushed by Microsoft?"

"Is good design correlated to success?"

A few weeks ago, these kids would have been overawed and intimidated by having such close contact with a real, live venture capitalist. Reiss-or any other VC--would have been the Willy Wonka of their world, the maestro whose attention could only be had if you were lucky enough to find yourself in possession of a golden ticket. Now they seemed to look at him with the polite curiosity of fellow dinner party guests; they were fascinated, and eager to learn, but they knew they didn't need him. Or anyone else, for that matter.

Six weeks later, as the Summer Founder's Program drew to an end, one of the eight teams actually turned down a lucrative acquisition offer. After all, who wants a golden ticket when you can write your own?

*

By summer's end, not one of the eight teams in the Summer Founder's Program had failed or folded; and it was clear to everyone who had participated that there was more going on here than luck.

"Four of the eight companies have closed deals for further capital or are in the process of doing so," Graham reported happily at the close of the program, before leaving for a two-week vacation. "Two more are continuing on having decided that they don't need further investment at this time. And the other two groups are going back to school, but will probably continue working on the side."

There are no statistics for the success or failure rate of angel investments, but if one takes the commonly accepted benchmarks for success in the technology industry-that only one in ten venture capital-funded startups (that's a minimum of $3 million and a lot of due diligence)--then Y Combinator has done exceptionally well for itself: for a total investment of less than half a million, it now has an equity stake in six viable companies, the majority of which have already attracted outside investment after being in business for only a few weeks.

So what was the essential ingredient, was it the seed money, the advice, the camaraderie, the glop, or all of the above?

It's a good bet that it's not the money. Firecrawler, Inc., was started by the most world-wise crew, a group of Cornell graduates in their mid-twenties (including Lisa Ong the only woman in the program). Ong and Phil Yuen both left jobs as programmers at Microsoft to come to Cambridge this summer, where Phil's brother Gerald joined them. Their product was a subscription-based software service that crawls a client's company website looking for errors and inconsistencies. Six weeks into development, they had already bagged three clients, including a law firm and an investment bank (Livingston's investment banking connections helped them out here). By program's end, they were doing enough business to be self-sustaining.

Phil had absorbed Graham's number one teaching: "After you sit through a few of these dinners you learn that these guys' favorite word is 'frugal.'" And frugality has paid off: he had, in effect, taken a startup company from launch to market in eight weeks, with an initial investment of $18,000.

Phil had no regrets about leaving Microsoft. "I think I'm an entrepreneur at heart," he said. "I started my first company in my sophomore year at Cornell." That was during the boom. When the bubble burst, he went back to school for his master's degree, then took the "safe" job in Redmond.

Phil readily admitted that he didn't come to Cambridge for the capital, but for the environment, and to learn from the Y Combinator founders. The team members could have scraped together $6000 each any time they wanted to, and they could have quit their jobs and taken the same chance on their own. But they didn't. Something about the opportunity tipped the balance between dreaming about a company and taking the shot.

It seems that not every prospective entrepreneur is scared off by risk. Sometimes it's just the vast amorphousness of the endeavor that keeps capable people from starting their own companies--the inscrutability of government regulations, the mysteries of money, and the sheer loneliness of the endeavor. Starting your own company may not be any easier than getting into MIT, but at least there is a body of conventional wisdom and a set of steps for getting into a good school, even if it's up to you to execute them well enough to get in. What Y Combinator wants to do is create a similarly comforting and easily undertaken set of steps for another kind of brilliant high achiever to follow.

In that sense, Y Combinator is more than an investment group; it's a for-profit graduate program, a place where those who can do, teach. For if there's a void in the funding system for startups, there's an equally gaping one in education: business schools teach management, not entrepreneurialism, and there's nowhere else aspiring entrepreneurs can go to learn the particular skills that were on offer in Cambridge this summer. At the end of the program, some might emerge with viable companies. Others would leave with nothing more than a "learning experience." But there's one thing no one would leave with, and that's a student loan to pay off.

Graham and partners are careful to portray the Founders Program as an adjunct to traditional schooling. This subject of whether or not some kids might decide not to go back to school the next year made Livingston wince slightly. "We don't want to be in the business of luring kids away from finishing their education," she said carefully.

By the end of October, Y Combinator had received an avalanche of applications for the Winter Founders Program, to start in January, and had announced a one-day Startup School. "We were going to call it the "Day Business School," said Graham, "but then we realized that not a single thing we have to teach is anything you'd ever learn in business school."

For the Summer Founders program is "grad school" only in metaphor. In fact, the program is itself a startup company, and its primary mission is to turn a profit. If it succeeds in doing so, it will probably be because it offers a solution to one of the biggest problems in mainstream industry today: how to create a model that systematically sustains and fosters innovation.