There's a school of thought that says entrepreneurs learn best from their own mistakes. That they must fail - and fail and fail - before they succeed. Noam Wasserman comes from a different school of thought (Harvard Business School), and he suggests that entrepreneurs instead learn from the mistakes of other entrepreneurs.

Problem is, they haven't had a handy database of common errors to refer to. Until now. Wasserman has just written one, titled The Founder's Dilemmas: Anticipating and Avoiding the Pitfalls that Can Sink a Startup.

"There's a craving for a roadmap that entrepreneurs have never had, to be able to anticipate things that in the past would have taken a serial entrepreneur with the pattern recognition to see that decision X will lead to outcome Y," says Wasserman, who's an associate professor of business administration at Harvard.

Wasserman has spent decades researching startups and why they succeed or fail. He's compiled a database of experiences from 10,000 company founders, and his new book mines the data for lessons, with stories from entrepreneurs like Evan Williams of Twitter and Tim Westergren of Pandora.

We talked to Wasserman about the most frequent causes of startup failure.

People problems. One of the surprises of Wasserman's research is that it's everyday choices that sink the majority of startups. "The most common decisions are in fact the ones most fraught with peril," he says.

Like who to work with.

Wasserman discovered that most failed startups are done in by people problems, such as picking the wrong cofounder. "The most common decision is to cofound with someone you know socially," he says. "But that's the least stable of all teams, because you'll likely be cofounding with someone who's more similar to you than they should be."

Cofounders who come from the same background have too many of the same skills, which leaves holes in important places.

And then there are the emotional issues. It's harder to be blunt and make the tough decisions about equity and responsibility when you're dealing with friends, Wasserman says. "You could be far less likely to handle the tension-filled discussions and tackle the elephant in the room until these issues blow up in your face."

To avoid that mess, start your company with someone you know professionally. Wasserman's research shows the most stable cofounder teams are those formed from prior professional relationships.

Equity splits. Wasserman says 73% of the teams in his dataset divided the equity within a month of founding. Too early.

"This is when uncertainties for the venture are at their highest, when you don't really know what your business strategy will ultimately be or the roles you'll ultimately be playing. Yet founders are splitting equity then and setting it in stone and it causes major tensions later."

He says it's OK to divide equity early - if you build dynamic elements into your agreement to allow for flexibility in the future.

Investor relations. Most founders take money from family and friends. "But you're playing with fire when you take money from Aunt Sally," Wasserman says, "and if you play with fire, you have to build some firewalls."

He suggests writing a prenup and discussing worst-case scenarios up front. "And thinking hard, as founder, that if Aunt Sally is the only one willing to invest in your business, maybe that's telling you something."

Of course, professional investors come with their own concerns. They add value, but they also bring risk - control risk. "You're opening up the possibility of being fired from your baby," Wasserman says. "By the time founders in my dataset raised their C round of funding, 52% had been replaced as CEOs and, of those, three-quarters were fired by the board."

Not an ideal exit. Then again, if you are escorted to the door with your laptop and your cardboard box, you can always go start another company. Just remember to learn from your mistakes.

Image courtesy of Shutterstock.