Who determines the upgrade strategy in your organization when it comes to software? If your answer is anything but the vendor or the market, I suggest that you rethink your answer. Although I would like to say that we have complete choice in how and when we upgrade, the truth of the matter is that as competition has dwindled in the market place, so has our ability to drive the decision to upgrade.
Let's take desktop operating systems as an example. You will not be able to purchase Microsoft Windows XP after January 31, 2009 from a systems builder and it will no longer be available to buy from OEM or retail after June 30th 2008. The only exception to this is if you participate in a volume licensing agreement in which case you can exercise your right to down grade - however Microsoft doesn't care because you have paid for the upgrade anyway in your licensing agreement.
This policy apparently has at least 64,000 people upset because that is how many people have signed Infoworld's petition to extend the life of XP. While many of the signees of the petition state that Vista is not worth the trouble to upgrade, a recent announcement from MS regarding their earnings state that Vista sales currently and those in the pipeline appear strong. Thus there are enough people out there that either agree that Vista is the better product or have given in to the inevitability that you cannot fight the juggernaut that is Microsoft.
I personally do not feel that Vista is a huge improvement over XP and tend to believe that Vista sales are being driven by sheer resignation - after all, what choices are there? At the home level, you can choose to stay with XP as long as there is not a new game coming in the next 6 to 12 months that you want to play that requires DirectX 10. If you do - well you will have to make that decision to "do without" or bite the bullet and upgrade. Some will argue that choosing to do without or upgrade is a conscious decision but I think that decision making when you have little alternative is not much of a choice.
At home, you could decide to stick with XP or switch out to Linux or buy a Mac - but that is also a decision to marginalize yourself when it comes to alternatives. While I love both operating systems, there will be plenty of mainstream products that you are cutting yourself out of because of that decision - particularly if you are a gamer. If you stick strictly to music, browsing, and productivity - you are less marginalized for sure - but there is still some cost when it comes to flexibility.
At the corporate level, given my argument above, you might think that the decision is not so cut and dry then. After all, there are equivalent productivity tools to run on an alternative OS than Microsoft. At least there is in my opinion. However making the decision to dump Microsoft's OS in the corporate environment is not a "safe" decision. These days, IT departments often describe themselves as a "Microsoft shop" and going against the Microsoft strategy would take a wholesale change in the culture of an organization. Besides, as the old saying goes - no one gets fired for choosing ____-it used to be IBM but now can be safely replaced with Microsoft. Add in the fact that many organizations participate in volume licensing agreements and thus change is only considered at renewal time - people tend to do what is easiest - just sign up again.
This is sad actually, because now would be the ideal time to make the argument for a change to something else. It is no secret that there will be a significant cost in migrating to Vista both in end user training and with the replacement of machines that can't run Vista or in the replacement of legacy equipment that Vista will not support. Add in Office 2007 which is wildly different than previous versions and you have a significant training effort to be able to do precisely what you used to do only with different equipment. This being the case, could you not make the argument that if you are going to have to endure pain, why not do so in an effort to gain more flexibility?
Actually, I think large organizations that are wrapped up in licensing agreements are more likely to pay for the upgrade and skip Vista by exercising their right to downgrade and wait for Windows 7 in hopes of something better. Again, this is a decision only in mind because the upgrade has been mostly paid for anyway - excluding the hardware and training.
Just to make sure that I am not picking on Microsoft, I will point out that this phenomenon works the same in the database space and the ERP arena as well. Consolidations in those markets have made certain that Oracle and SAP have a say in your upgrade strategy as well.
In the end, unless one works very hard to stay vendor agnostic, (and it is hard work) our ability to determine when change is needed has been dramatically reduced over the years. The decision today is not when you want to upgrade but to which company you are going to hitch your boat to and take the ride.