Upgrading to Windows 7 isn't Cheap

A recent Gartner report showed what many of us already knew: Moving to Windows 7 from XP is expensive.

I've upgraded old XP PCs to Windows 7 and I've bought new PCs equipped with Windows 7. Either way I've found that it's not cheap. Now, Gartner, the research company, has worked out just how expensive upgrading to Windows 7 can be for enterprises. In one word, moving to Windows 7 can be described in one word: "Ow!"

Charles Smulders, Gartner's managing VP believes that "Corporate IT departments typically prefer to migrate PC operating systems (OSs) via hardware attrition, which means bringing in the new OS as they replace hardware through a normal refresh cycle. Microsoft will support Windows XP for four more years. With most migrations not starting until the fourth quarter of 2010 at the earliest, and PC hardware replacement cycles typically running at four to five years, most organizations will not be able to migrate to Windows 7 through usual planned hardware refresh before support for Windows XP ends."

In other words, Smulders claims you're going to update faster than you had budgeted for. That presumes, of course, that you've budgeted at all for upgrading your desktops. Times are hard and I know many companies where the 'upgrade' plan is to run PCs until they break.

That said, Gartner still thinks businesses have three Windows 7 upgrade options. These are:

Accelerate PC Replacement Plans

The plus side with this plan is that you know these machines will work with Windows 7. I recommend buying new Windows 7 PCs to replace older XP PCs myself.

It's also believe it or not, probably the cheaper way to do it. According to Gartner, "Assuming a 10,000 PC environment, where all PCs are replaced, Gartner estimates that the migration cost per PC will be between $1,205 and $1,999, depending on how well-managed the environment is. While the overall cost to migrate is lower than other scenarios, the down side is that the capital costs account for about 60 percent of the total replacement cost, so the capital budget will be larger than in the upgrade case." That sounds about right to me.

Upgrade Installed PCs

Here Gartner states that "Using existing PCs will reduce the capital costs of migration, but will not reduce the labor costs of migration. Assuming the same setup as above - a 10,000 PC environment, where all PCs are upgraded - the migration cost per PC will be between $1,274 and $2,069, depending on how well-managed the PC environment is. This assumes that 25 percent of the machines will need a hardware upgrade to run the OS."

The problem here is that "While the capital costs are reduced in this case, upgrading an installed PC simply postpones the inevitable replacement for two to three years. Users will need to be migrated twice, rather than once, during a four-year period."

I disagree with this. From my experiences with upgrading XP PCs to Windows 7, the labor costs are going to be higher than Gartner's estimates. Indeed, I think it will be a lot more expensive. I've think $1,500 to $2,500 is far more likely.

Gartner also recommends a Partial Migration. This isn't what it sounds like. This isn't moving just some PCs to Windows 7 while leaving others.

Instead, Gartner suggests that "For task workers, such as data-entry roles (these account for about 15 percent of the population in a typical organization), migrating from a PC to a hosted virtual desktop (HVD) environment is an alternative to PC migration. It would potentially speed up deployment, because it is one image deployed centrally. However, an HVD does not solve the budget issues, because of the incremental cost of the data center and network infrastructure needed to run an HVD. Also, it does not solve the IT support staff issue, since they will be involved in the HVD rollout."

What Gartner doesn't suggest, is that is no reason to spend cash on Windows licenses or an HVD. For task workers an inexpensive or free Linux desktop is a perfect fit. There should be no Windows specific applications at this level, the capital costs are low, there's no need for a hardware upgrade, and, unlike Windows, there's no need for expensive PC security software. You don't need to be a Linux fan to see that this makes hard business sense.

Another option that Gartner doesn't explore is just sticking with XP to the bitter end. After all, that may be a while in coming. Microsoft recently extended XP's life. 2016 sounds like a fine year to me to look into another operating system for my desktops.

These last two options of mine also have the real advantage of being dirt-cheap. This is not a small consideration. Steve Kleynhans, Gartner 's research VP said in a statement that "Based on an accelerated upgrade, we expect that the proportion of the budget spent on PCs will need to increase between 20 percent as a best-case scenario and 60 percent at worst in 2011 and 2012. Assuming that PCs account for 15 percent of a typical IT budget, this means that this percentage will increase to 18 percent (best case) and 24 percent (worst case) which could have a profound effect on IT spending and on funding for associated projects during both those years."

And, where I ask, do you think you're going to get the money for that? Exactly. You don't have it do you? I think your real IT budget plan for PCs for the coming years is to deploy Linux on the low-end or at higher desktops if you have Linux-savvy staff, and stick with XP for everything else. After all, it's not like you have the budget for anything else do you?

Insider: How the basic tech behind the Internet works
Join the discussion
Be the first to comment on this article. Our Commenting Policies