Failure is data

A popular debate in the tech world is whether failure is good or bad. Eric Schmidt reminded people that at Google, “we celebrate our failures.” Taking a contrary position was Jason Fried from 37Signals, who wrote back in 2007 that “I’ve never understood Silicon Valley’s obsession with failure.”

NPR’s Melissa Block recently visited Silicon Valley to talk about failure, interviewing (among others) Joe Kraus. Joe is who hired me to run Blogger several years ago, and is who brought me to Google Ventures last fall. The whole piece is worth a listen, but I found Joe’s comments worth highlighting:

In my mind, the ones who have no fear of failure are merely the dreamers, and the dreamers don’t build great companies. The people that thread the line between vision and being able to execute and having this healthy fear of failing that drives them — not paralyzes them, but drives them — to be more persistent, to work harder than the next person, that’s a magic formula.

This is the important distinction. When I say I celebrate failure, it’s because I see failure as data. Data itself is neither good nor bad: it’s just data. But what you do with that data – how you learn from it, how you apply it to future decisions – that’s why you “celebrate” failure. Failure in this sense is just a way to help inform future success, not an outcome to be celebrated.

You don’t start something assuming you’ll fail. And you don’t pursue an idea without some awareness that it could fail. But fear of failure shouldn’t stop you – as Joe notes, it should motivate you.

I was fortunate to address the 2009 graduating class at my alma mater, Richmond Law. In that address, I said this about failure:

When you think differently, accidents happen. Failures are unavoidable. But accidents aren’t always failures, nor are failures always without value. When it’s OK to fail, success becomes possible, and, through experimentation, the next step in the path presents itself.