So innovation comes from the employees often. One example is Gmail, something I worked on. That came from one employee, that thought that he could make a much better mail client. This seemed crazy, but he worked on it in his 20% spare time, and convinced some other people, such as myself... we convinced more people, such as Larry and Sergey, and I believe we made one of the best mail clients in the world.

Indeed the Google jobs page today states, We offer our engineers 20-percent time so that theyre free to work on what theyre really passionate about. And Google in their original Gmail press release told the story Joseph repeats, saying The idea that there could be a better way to handle email caught the attention of a Google engineer who thought it might be a good 20 percent time project. (Google requires engineers to spend a day a week on projects that interest them, unrelated to their day jobs).

On the other hand, ex-Google employee Paul Buchheit at Friendfeed comments:

Oh no, now even Joseph is repeating the Gmail in 20% time myth! ... It wasnt a 20% project " it was my regular project. It predates the 20% time rule, in fact (though I always had side projects anyway).

Someone named Hans Cardinal, who claims to be an ex-Google employee (which may or may not be correct), put it more drastically in comments in relation to another story at Valleywag: Google recruiters are out of touch. This 20% project concept is a thing of the past yet they still promote it as if it exists. Recent survey (Googlegeist) shows that most people dont even have a 5% project since theyre working overtime on their primary project. Whether or not this is more closer to the truth is hard to tell from the outside; perhaps we need to take it with a grain of salt just like the commercials.

In the meantime, Joseph replied to Paul at Friendfeed " perhaps he didnt mean it as literal as the video might make you believe. Hes writing:

Dude, dont make me look at the video again. They told me not to move my arms, made it dire. Anyway, I still claim gmail is a prime example. 20% was never a 20% rule, it is more the -mindset- that you can muck around and do stuff. You hacking up the usenet code for a day to do your personal mail fits that.

Update: Googles Mihai Parparita says, A little-known fact is that Google Reader started out as a 20% project (or if you wish, not strictly during 20% time, but definitely as a side-project/bottom-up kind of thing, like Gmail).
Paul Buchheit in a follow-up to me says, 20% is important as a concept because it gives license to work on things that other people dont think are important. Various people inside Google would periodically try to organize or coordinate (or approve!) 20% time/projects, and thats a mistake because it runs counter to the value of 20% time. Elsewhere he clarifies that he thinks the 20% time idea is very valuable. And Googles Matt Cutts comments that Google is better than most companies at supporting engineers who want to try out a new idea or a side project.

Del.icio.us

Digg

Reddit

Furl