Trusting Data, Not Intuition

Yeah and when software that vendors thought was great was evaluated in the real world it was rubbish.
... are evaluated through controlled experiments, less than 50 percent actually work out...
50% that's a pretty good success rate. If we gave up thinking because only 50% of our ideas were good we'de soon be condemned to oblivion.

Oh sorry, just realised it's at Microsoft... I shouldn't be expecting joined up thinking.
Del
(oh I'm a baaad cynical kitty, I shall go and castigate myself on the naughty step)

My experience is that many projects are ego based. The purpose is to get their champions more status, overseas trips, more staff, improved profile or to thwart opponents. If the project works to the companies benefit it's a bonus.

I know I have to estimate a "return on investment" for each project I propose, luckily no one ever goes back and checks if my guess was any where accurate.

Although, given it's accountants who run the system, real world results aren't that important. Ffej

If only 50% of your ideas pass the scrutiny of real world experimentation, then double the number of ideas you are testing...