Eliminate The Development Machine to Increase Innovation

In my quest for knowledge and insight about how great companies are managed, I've been reading "The Google Way" one of the, I'm sure, myriad of books that purports to explain all of the things that make Google amazing.

One of this weekend's insights was the following:

A lot of good ideas (and a few great ones) come from rank-and-file employees.

In a "traditional" organization, many of these ideas are overlooked. Great ideas are supposed to be generated by people tasked with dreaming up the products of the future, the select few "tinkerers" in advanced engineering or some elite department of marketing. These ideas are then evaluated in committee, picked apart, perhaps improved, and documented. The few ideas that survive this process are then given to a marketing organization to see if a business case can be made for these projects.

In an organization like Google, the leadership believes that ideas are generated everywhere. Idea generation is encouraged and expected of all employees. The most obvious way this philosophy shows up is in the policy that engineers may devote 20% of their time to working on their own projects. Some of these projects have died, but some became Google's better known products. Would gMail or Google Maps have died in committee at a traditional organization? Likely; neither has a direct connection to Google's core business of search.

In my organization, most of the good process improvements (by which I mean changes in the way code is developed, not institutionalized processes) came from rogue engineers who were spending time working with tools beneath the radar of management. Software engineering at my division is special in that the developers are given a lot more freedom than in other parts of the company and it shows; many other development centers come to us for advice and are interested in adopting our best practices. Unfortunately, the "product development process" that we use ensures that no product ideas that come from the rank-and-file engineers will see the light of day . . . and I believe that will ultimately cost us as talent goes elsewhere.

How does your organization work? Do you feel like you are encouraged to explore newer and better ways of working? Have any new products come out of your rank-and-file?