I've just come from a 90 minute technical presentation from a vendor. It occurred to me during my normal, incessant questioning of the presenter that what I really wanted to hear is what the product is NOT. That is much more helpful than the typical "scalable, enterprise-ready, application integration" blah blah drivel.
Only in contrast to similar products can we really tell what something is.
Of course, as a sales person, I want to think that my product can do/be anything. <g>
This is semi-related to the idea of the domain diagram early in a system's life: tell me what it is not.
Also to the very important "Scope" section of an engagement definition -- what are we NOT agreeing to do.
And something I remember from a brush with Developmental Psychology -- the importance of a child discovering the boundaries to herself -- "My hand, that is part of me, but the cat is not."
Sometimes negative approach helps to make decision. Reducing scope is good when there is lack of time. Though I prefer choosing best, not avoiding the worst.