It also depends on you having a crack IT team who can get it right the first time. No months of stuff crashing or having a horrible interface that impedes work or being intolerably slow. You have to be confident enough to push back against the exec who says "oh, we'll just hire $massive_consultancy_firm to do it" and takes years to roll out a half baked hackjob on top of their stupidly expensive CMS.

That sort of nepotism can be a problem. It's why Amazon's search sucks so bad. In the book 'The Everything Store', it tells the story of when a team at Amazon responsible for something else saw how bad search was and decided to fix it. So they built a prototype with Elasticsearch and whatnot. It worked great. The dude at the head of the search team, good friends with Bezos, got all territorial and angry. Bezos solution? Have a test between the two solutions to determine which was best. Bezos idiocy? He let the head of the search team decide who won. So, no surprise, the head of the search team said the search team won. So the improved search was shut down and the garbage search we all have to deal with regularly got to stay.

Once you've got that kind of thing going on, only retirements and deaths can effect good change at an organization.

It is, in the sense that I wouldn't expect it to happen, at least to me. But it's not unrealistic that some CEO would actually listen to you about the matter and put his foot down to make it happen. These are not matters difficult to comprenhend, my guess is that people use to take IT stuff for granted and that's it. I've also known guys who didn't give a fuck when they saw upper management screw up tech-wise so as not to get into arguments. That also cripples healthy feedback.