I now write on Medium here. You can find some of my old essays below.

Learning from Failure is Hard, Learning from Success is Harder

Unqualified lessons are hard to wring from startups. It's difficult to really understand something without doing it, and no two people who do it have identical experiences. That said, I've seen two schools of thought on how to learn from others' startup experiences: 1) It's best to learn from success. Stories of failures have no prescriptive advice on how to accomplish something. Stories of success, on the other hand, offer models to copy.

2) It's best to learn from failure. Success is fairly random, whereas failure usually happens for distinct reasons.

First, anyone thinking of starting a company should take lessons from wherever they're available -- successes, failures or those yet to be determined. But given a choice, the best lessons come from failure.

I've stopped thinking of a startup as a discrete success or failure. Rather, a company is a series of small (or big) successes and failures, one side of which eventually overwhelms the other. Some failures are singular and crushing -- taking on a bad co-founder, for instance. Others are subtle and can sit under the radar for years, even in companies that are otherwise successful. Poor initial equity distribution or choice of corporate structure, for instance.

That said, most companies aren't killed by one overwhelmingly bad decision. They're killed by dozens of bad decisions that pile up and stick the company in an unrecoverable morass. These small, could've-gone-better decisions offer the best startup learnings, are the exact lessons covered up by success.

Take one of the greatest entrepreneurial success stories of our generation -- Facebook. Clearly, Facebook did some things right. But they also did a whole lot of things wrong, and it's hard for anyone to determine how much more Facebook could've screwed up before they would have doomed their chances to win the social networking wars. It's entirely reasonable that they could've blown any number of decisions and still come out on top.

On the flip side, copying success is tricky. Let's look at Facebook again. The most difficult thing about copying Facebook's success (or any company's success) is not deriving the factors that led to their success, but figuring out the level on which they should be copied. For instance:

Level 1: Facebook was successful, so I should copy what they created: a new social network. Good luck with that one.

Level 2: Facebook was successful, so I should copy their operational or strategic decisions for my own idea. As I pointed out above, it's not obvious that Facebook made great operational or strategic decisions.

Level 3: Facebook was successful, so I should copy their methodology of thinking about ideas or products. This sounds high-level enough, but how does one execute on it? Does anyone know what goes on inside Zuckerberg's brain? More practically, are those thoughts relevant to your product as opposed to a walled garden social network?

On the theme of translating theory into practice, I helped create Founders at FAIL, a forum for entrepreneurs to talk about their failures. Think Founders at Work, but all the stories have unhappy endings. I'll be speaking there (along with GoCrossCampus founder Matthew Brimer) on August 18th.