Fixing scope up front says win-lose to me. It’s not the basis for a happy relationship. Everyone is best of friends while things appear to be tickety-boo, but that’s just an illusion. The moment tickety-boo goes bang the contract kicks in to identify who can be blamed and penalized. But I digress.
I seriously doubt it’s possible to know everything up front and to expect success on that basis is, well, wishful thinking at best.
As people making things for users, we don’t have all our ideas at the same time, and we don’t have all our good ideas before we begin doing something. Also, things change, Murphy’s Law has a say, shit happens. When it does, it just might cause us to seriously rethink what we spent time and money figuring out at the start. The simple fact is that we learn more as we go, and if we’re to improve the odds of succeeding we need to remain receptive and responsive to what we learn.
As users of things, I doubt we truly know what we want. We need a chance to know what a product is capable of doing and how we might cope with it, before we can define what our real requirements are. We can only do this by using the bloody thing.
Of course there’s got to be something up front to provide context and initial direction. The skill is to do just enough, then have at it like a Whirling Dervish, iterating furiously and testing continuously to discover the unknowns that need to be known, learn the right thing to make and the right way to make it.
