Not so fast. In the Harvard Business Review, Chris Trimble challenges the convention that Google is able to pervasively implement this policy. And that it's even a good idea to begin with.
The problem, according to Trimble, is that innovation requires both good ideas and a significant investment in execution. 20 percent of someone's time might yield some interesting concepts, but even a company as vast and resourceful as Google can't make a reasonable investment in executing on all the good ideas that arise. You either get half-baked products like Google Wave, or a lot of stuff left on the floor, completely unimplemented. Either approach is a waste of resources and bad for the company.
Personally, I see this problem firsthand -- innovative thinking is frequently encouraged at many companies I interact with, but it's clear that there's no way that management is prepared to invest in executing on any of those ideas. I suspect managers idly hope someone will come up with some awesome, innovative idea that requires no capital, no resources, and no engineering support to implement. Because that's frequently the only kind of ideas there's any practical way to deliver.
What's your experience with innovation goals? Sound off in the comments (in an innovative way).