Google has garnered a lot of attention and some success with its "20% time" idea, which enables every engineer to spend one day a week working on projects that don't fit in their job description. In my observation, just announcing that every engineer is expected to spend a certain fraction of their time on innovative ideas won't magically lead to innovation. Plus, it's very hard to implement the 20% time model at a startup: most startups just don't have the luxury of 20% excess engineering capacity.
At my (startup) company Kosmix, we take a somewhat different approach to create a culture of innovation, which I described to Taylor Buley of Forbes in a recent video interview. I think the video is terrific, and encourage you to watch it (it's also embedded at the bottom of this post), but there's only so much that can be said in a 90-second video. So I collected together some of my thoughts into this blog post.At Kosmix, we don't specify a set fraction of time for people to spend on new ideas. Instead, we have focused on creating a culture that engenders new ideas and rewards innovators, encouraging them to tackle new projects above and beyond their 100% contribution to mainline company execution. The three key building blocks that we've used to create a culture of innovation at Kosmix are Team, Environment, and Incentives.
Team. It always starts with the people. At Kosmix we are fortunate to have a team of rock-star Computer Science graduates from top universities; it's hard to throw a brick without hitting a PhD. Since CS skills are taken for granted, the interview process emphasizes creativity, problem-solving skills, and teamwork skills.
Very importantly, many Kosmixers are multi-dimensional people with interests and passions that extend well beyond work. For example, one of our Operations gurus has a deep interesting in (hold your breath) knitting, and runs knitting classes at work (they're called Knitting Knights). Our office manager also happens to teach Art History.
Environment. There's something about the graduate school environment that seems to bring out great ideas. Many of the great technology companies (e.g.,Yahoo and Google) have been created by graduate students. We have strived to maintain a grad school environment at Kosmix. Wall around and you'll hear plenty of heated hallway discussions and intellectual free for alls; nerf gun fights erupt over details of relevance algorithms.
When I was a grad student, I used to get ideas for whole new lines of research by attending talks by other students and faculty. The Infolab, the research group I was part of Stanford, has a tradition of Friday lunches where a student leads a discussion on their ongoing work. We have copied this model at Kosmix: every Friday, we have a communal lunch gathering, and a Kosmixer leads a discussion -- either on something cool they've been working on, or on some topic that's just cool but completely unrelated to Kosmix -- such as muscle cars, alternative fuels, or astronomy.
Incentives. Given the right environment, the next piece is incentives for people to go above and beyond the call of duty. At Kosmix the biggest reward is peer recognition through a system of awards:
- The Kosmix Kreed award is peer recognition at its purest.
Any Kosmixer can nominate any other for doing something interesting and
inventive that helps Kosmix users, or for going out of their way to
help out another team or person on working on a different project.
Giving this award is as easy as sending an email to HR, with a clear
description of the achievement that merits the award.
- The Just Do It! award is given by management, and recognizes an individual who did a substantial project that goes above and beyond their job description. We stole this idea from Amazon.com, where some of us used to work. For example, one recent awardee dreamt up, designed, and implemented the feature that allows users to customize the Kosmix homepage, without any directive from management. Another implemented the ability to edit any topic page on Kosmix.
One of the big successes of the Kosmix culture of innovation has been Meehive. A while back, a Kosmix developer thought it would be cool to take Kosmix's core categorization technology and apply it to the problem of filtering news and blogs. He worked on it for a bit to create a first version, which convinced management that this was important enough to create a full team around. We then staffed an official Kosmix project to create Meehive, a personalized newspaper, which we launched last month. You can specify your interests very easily (I have over 40, including technology and cricket), and Meehive scours thousands of newspapers and millions of blogs to create your own personalized newspaper. Early adopters love Meehive; I now use it as my main source of news every day. Check out what people are saying about Meehive on Twitter.
Oh, and by the way, the most recent Just Do It! Award went to a developer on the Meehive team who took it upon himself to create the Meehive iPhone app. It's now rising in popularity among News applications in the app store, and has been a bigger success than any of us imagined. Best of all, no told the developer to do it.