SiliconBeat

With millions of users and more joining every second, LinkedIn has become a data nerd’s delight. And it allows for intriguing experiments like the one recently conducted by Monica Rogati, LinkedIn’ senior data scientist.

Rogati was curious: What could that rich set of data tell her about the factors that go into making an entrepreneur? Or, as she put it in a study released today, “Sequencing the Startup DNA.”

If there was a big surprise for her, it was this:

“Geography matters, even if you like to think it doesn’t,” Rogati said. “Even if you like to think starting a company is democratic and the world is flat.”

Rogati basically took about 10,000 profiles of people who had started companies in some fashion. Interesting to note that only about 2 percent of those people go on to start another company, becoming so-called serial entrepreneurs.

But looking at geography, the most likely place someone will start a company is San Francisco. Indeed, someone is twice as likely to start a company in SF as they are in New York. And in turn, they are twice as likely to start a company in NY as they are in Boston.

The data on age is also interesting: 40 percent of founders were between 30 and 39 years old; 20 percent between 40 and 49; only 34 percent between 20 and 29. That means that over 40 percent were over 30, which challenges the conventional wisdom about startups being a game for the young.

The other data point likely to get tongues wagging is the schools where founders are mostly likely to come from. Number one is Stanford’s Graduate School of Business, followed by Harvard’s biz school, Berkeley’s Haas School of Business, MIT Sloan, and Tuck School of Business at Dartmouth.

There’s other data in there, of course. Check it out and post any other thoughts below.