The innovation gamble


When it was founded, engineers at Google were allowed to take 20% of their time to work on whatever they wanted. Gmail was originally a 20% time project. A slightly modified version of that policy still exists today.

I remember a long time ago reading an article critical of this policy, or maybe critical of Google generally, or possibly supportive of Google and its 20% time policy but with a particularly pithy remark in it. What I’m trying to say is that all I remember about this article is that it suggested most people spend their 20% time building model ships and doodling instead of making products that are of actual value to the company.

I wasn’t sure why that stuck out, but it did. I think I remember agreeing with the sentiment at the time, but now it strikes me as short-sighted.

Creativity and innovation are bizzarre, non-linear processes. You can’t sit someone down and tell them to innovate. If you want cool new ideas, you have to foster an environment where they can happen organically, and then cross your fingers. And you can never predict in advance exactly how or when it’s going to happen.

Your best chance, if you want to build something crazy and new, looks something like this:

  • Hire the best people in the field. As many as you can get.
  • Give them as much freedom as you can afford to give them.
  • Support them and their crazy ideas any way you can.

With just one person working this way, odds are you’ll never get a great, market-defining innovation. Your odds go up slightly with every person you add. Innovation-wise, every person you add is like another lottery ticket. Adding more people brings you closer and closer to certainty that you’ll end up with something cool.

It’s a guarantee that they won’t all invent Gmail. You’ll get a thousand duds for every killer appl. The point is, all of those model ships that got built are the table stakes - you must let them happen. The innovation can’t happen without them.