Things like who you promote, how you promote, whether to have an engineering ladder, how to reward engineers, etc., have a deep and lasting impact on who does well in the organization, and who gets pushed out. That lasting impact in turn impacts how decisions get made and who gets listened to. It's not surprising, for instance, that Yahoo! had a deep blind spot about search and thought it was a media company, or that Microsoft was clueless about why Vista was going to be a disaster, and that Google in turn, ignored social networks until this year. If you trace all those blind spots you eventually come to the root cause, which is the way the organization was built, what got valued, and what got discarded.
This is not necessarily a bad thing: if you look at successful organizations, they all do better when they focus, rather than spread themselves too thin. In the case of Microsoft and Yahoo, though, that focus cost them billions, and despite Microsoft spending billions it hasn't fixed its internet blind spot, and Yahoo doesn't look like it's going to regain the eminent position it once had as a portal. It is too early to tell whether Google's turnaround on social media is happening early enough or quickly enough (i.e., we don't know whether Google is Microsoft circa 1996, or whether it's Microsoft Circa 2001).
In all cases, I believe that the problem is lack of independent thinking at the top levels of the organization, and the problems had the root cause well before the disaster happened. (i.e., the problem at Yahoo happened with Terry Semel's hire, not with Semel's decision to buy Overture, and the problem at Microsoft happened with the decision to ignore the internet after killing off Netscape when Brad Silverberg lost the political battle inside Microsoft and retired, not with the installation of Ballmer as CEO)
Please join in on the conversation if you have questions or have anything to add. I'm not as familiar with Microsoft/Yahoo as I am with Google (though I did meet Brad Silverberg once), so please let me know I'm wrong.