Geoffrey Moore has a blog (via Doc Searls) titled Dealing with Darwin, where he shares his viewpoint on business and innovation in a post titled "Beyond Innovation". Here is a paragraph, go read the rest.
To be valuable from a business point of view, innovation must provide a
deviation from the norm that creates differentiation in the company’s
offering which in turn leads to customer preference at the time of a
buying decision. In this light, the big challenge may not be coming up
with the initial deviation—there are usually lots of good ideas in
play. No, the real challenge is coming up with all the supporting innovations that reinforce the initial vector,
aligning all the other functions in your company to reengineer their
processes in such a way as to further accentuate the new value
proposition, thereby creating a sustainable differentiation that can
generate deep and lasting competitive advantage.
With great clarity, his focus is on usercentric differentiation based on aligning your offering around a singular value proposition that supports the core competency of the initiating organization. And he ends with these words,
People like to say that innovation has to bubble up from below. Maybe at the start. But lasting differentiation cascades down from above. Innovation management, in other words, is as much a part of the outcome as the initial spark of innovation itself.
4 comments so far
And additionally, I thought it was good to mention the concept Moore makes of ‘enhancement innovations’. I don’t know if it has been discussed here, but there can often be a lot of attention on single aspects of innovation that deliver consumer products in place of looking anwhere else.
RE>lasting differentiation cascades down from above.
Sounds like we need management innovation 
Gary Hamel says something similar:
http://www.businessweek.com/innovate/content/aug2005/id20050811_693230.htm
Damien, thanks for bringing that up, sort of like, if we find the platform on which to innovate, one platform should provide myriads of options and line extensions?
Victor, possible invitee? 
Moore is doing something much needed in discussions around innovation: he is admitting that there are different kinds and they must be treated differently.
This is something I’ve been working on a lot recently, and my latest blog post called Categories of Innovation discusses exactly this issue.
Where I use 4 categories Moore uses 14. My gut reaction is that for most people, 4 is easier to grasp than 14. However, since his book doesn’t come out till after Christmas I can’t tell if his model needs to go on a diet or mine needs a few cheeseburgers.