I'm a big believer in failing fast and learning. If you want to innovate, you plan a little bit, create a whole team, give them a wiki, foster a collaborative/transparent culture, set a short deadline, try something, measure it, and do it again until you are done.
I whole heartedly agree with this excerpt: I am an unabashed fan of MIT Media Lab guru Michael Schrage—particularly his book Serious Play. His principal axiom: "You can't be a serious innovator unless you are ready and able to play. 'Serious play' is not an oxymoron; it is the essence of innovation." And, in turn, the heart of his serious play is ... fast prototyping: "Effective prototyping may be the most valuable core competence an innovative organization can hope to have." His intriguing connection, which makes all the sense in the world to me, is that true innovation comes not from the idea per se, though it guides the work, but from the "reaction to the prototype." In fact, in a surprising number of cases (the majority?) the collective responses to a host of fast prototypes reshape the original idea beyond recognition—or lead one down an entirely new path.
A more succinct way to sum up how innovation works is Ready. Fire. Aim. Tom's article says he heard a Cadbury exec call his approach to product development. This phrase is often used as what not to do when it is exactly what you should do.
Here is a great quote:
"For C-sakes, quit drawing those f-ing maps and run some experiments, quick and dirty, and see if anything you are babbling on about actually works or makes the slightest bit of sense in the real world as we know it. And after you've done your real work, then you are welcome to write your 'complete theory of everything.'" (That was close to the actual script, minus many more f%^*s and about 25 minutes of elaboration.)