The other day I read an interesting piece on Eric "Astro" Teller, who heads the Google X department that does stuff, which seen from a perhaps slightly more earthbound view would be regarded as an R&D department for high-flying ideas.
The point is that when you run a department, which is limited (and who aren't these days?) with regards to time, manpower and other resources, it is imperative that one does what one can to kill ideas early - you simply must go to work with the mindset: "how do we kill our project today?"
It is an extreme interpretation of Linus Pauling's view on how to get good ideas: get a lot of ideas; hereafter you "simply" have to find out which ones should be discarded. And it is in excellent thread with so many other theories about failing quickly and celebrating mistakes, something at which we probably all should be better.
If there is just one good idea that you simply aren't able to kill, it would be worth it.
(Translated from Mordet på den gode idé, originally published April 28, 2016)