These massive project just don't work. 1) Pay consultants vast money to 'spec' out an insanely complex requirement. 2) Go through very long winded and expense t

These massive project just don't work. 1) Pay consultants vast money to 'spec' out an insanely complex requirement. 2) Go through very long winded and expense tender exercise, with the usual suspects - all expensive and with many failed projects under their belt 3) Company builds from scratch to the spec. 4) Spec is obviously incomplete, discover too late, company asks for lots more money to fill the gaps. 5) Rinse and repeat (minus the tender) You end up with something that has taken 10 years, and is delivering a 'requirement' that is 10 years old and now not relevant (even less relevant than the imperfect 10 year old spec). The only real way to do this is to break these projects up - either in scope or in size - ideally get something as off the shelf as possible and iterate it until it suits the needs, and then roll out wider. I.e. get well away from a waterfall process of massive scope. For example, instead of trying to put in place an NHS wide system (which then fails at great expense), get each individual Trust to go on their own with off the shelf suppliers or at least smaller, cheaper bespoke organisations. Smaller, more manageable and cheaper. You can define interoperability standards if that's needed; you can buy enterprise licenses if one supplier does end up being obviously better.