Topics >  Running a Business >  Success Stories >  Serial Entrepreneurs > 

The Start-up Guru

Kan's life attracted a few thousand fans and reams of press. But Kan soon noticed that instead of broadcasting from hat-cams, some users were interested in more traditional types of broadcasting. "People were e-mailing us saying, 'I want to broadcast a bike race or a talk show or a concert,' " Kan says. "We were like, 'OK.' " Kan stopped wearing the camera and focused on building a live video platform.

This kind of meandering path, Graham says, is encouraged at Y Combinator. "A lot of great companies started with different ideas," Graham says, noting that Steve Jobs's first plan for Apple was to sell do-it-yourself plans for building computers. "You need to listen to your users, figure out what they want, and do that." When founders are accepted into Y Combinator, they are given a gray T-shirt that says, "Make something people want." When a company sells, the founders get a black shirt that says, "I made something people want."

Despite having spent five years painting, Graham long ago put away his brushes. None of his work is on display in his home in Palo Alto, and he's none too eager to talk about matters of technique or style. But one thing painting taught him was the value of living frugally. "It taught me how to do cheap in a cool way," Graham says. Artists, Graham discovered, don't pretend to be rich; they live in sparsely decorated lofts and wear cool vintage clothes. "A start-up is that philosophy applied to business," he says.

Graham loves saying things like this. Investing in 145 start-ups so far -- and interviewing nearly 1,300 founders -- has taught him a lot. He has learned that start-ups with four or more founders tend to be less successful than those founded by smaller teams. "I think these people are afraid to do a start-up," he says. Another insight: If ownership stakes for founders are unequal, the start-up will probably get torn apart by infighting. "In theory," Graham says, "we ought to get to the point where we can just look at a start-up and be able to tell if it will succeed or fail."

Twenty-seven more companies will join Y Combinator in June, the program's biggest class yet. Earlier this year, Sequoia Capital, the venture capital firm that backed Google, agreed to give Graham $2 million to put to work. Graham hopes to use the money to expand Y Combinator, to fund 60 companies or more every year. I ask Graham why he is so intent on growing. Why does the world need so many little software companies? He looks at me as if I'm insane. "Imagine that instead of starting Google, Larry Page and Sergey Brin had taken jobs in some research lab," he says. "They would have written a little piece of an operating system that might not even get used and maybe some boring academic papers. Think of how much more they did for the world as start-up founders."

As the interview ends, we go inside his house and start to say our goodbyes. But before I can leave, Graham has something new to show me. It's a software program called EtherPad -- developed, of course, by a Y Combinator company. The program lets writers collaborate on a single document simultaneously. He thinks I might be able to use it back at Inc.'s offices.

"But what you really should do is start a start-up," he says. He stops and thinks for a moment. "You know what? You should do a news website for finance. You could be like Bloomberg, but you'd be just one person." His mind is going faster than his mouth. "What about Blogberg as a name? Is Blogberg.com taken?" Before I have a chance to respond, he flips open his MacBook to check, using a clever domain-name search tool developed by another Y Combinator founder, Beau Hartshorne.

"Damn," he says. The name is already registered. "Do you know anyone who can program?"

"Not really," I reply.

"Maybe someone from school?"

I recall that a number of founders have warned me that Graham has the habit of throwing out business ideas at a rapid clip, without really thinking through what he's saying. Still, says Kan, "you always come away feeling energized. You could be working on the most boring piece of software, and you talk to Paul and you think, Man, I'm excited to go back to work."

Indeed, there's something exhilarating about Graham's optimism, especially at a time when so many once-great companies are sitting on the verge of bankruptcy. Graham believes, deeply, that start-ups are the answer to the world's problems; that they are easy to make if you are determined enough and cheap enough; and that it's getting a lot easier to start one. We shake hands, and I leave, with a sudden desire to make something.

Max Chafkin, Inc.'s senior writer, wrote about Zappos.com for the May issue.