One of the more important features of an innovations program is failure. Innovations should not be low-risk (although they should be risk-managed, which is not the same things), which means that even in a well-managed scheme, some innovations projects will fail. Indeed some of them *should* fail – if this isn’t happening, then you’re either not reaching high enough to really make a difference, or your standards for what should and shouldn’t continue to be funded probably need to be adjusted. Innovations is about having the taste to choose a set of projects that might work, but also about having the sense to stop the ones that don’t. Which inevitably will disappoint several people: not just the people with the ideas that don’t get continued, but also the sponsors who put effort into backing those ideas too. And that disappointment needs to be managed, so that it doesn’t become corrosive to future projects, or to the idea of a working innovations program.