It Won't Hit in 2006

For starters, the new operting system won't be released until the second half of 2006. Only at that point will IT managers and other corporate customers receive final versions for evaluation. Also, only at that point will most IT professionals be able to receive any kind of formal training or cerfication on the new OS. What this means in practice is that no kind of corporate deployment will be practical before early 2007. Certainly, you might see a test machine here or there, mostly in IT departments, but nothing widespread will occur in the next year, and that's assuming Microsoft actually delivers on time.

It Doesn't Do Much That Matters

Second, many of the features the new operating systems will offer are really not all that important for corporate users, at least at this point, for a variety of reasons.

1) The "cool" factor of the new desktop graphics is unlikely to impress, or even be installed on most corporate desktops. For the most part it won't work all that well on older hardware and even some newer graphics hardware won't support it well. And corporate IT departments won't want to support it any more than they want to support a variety of wallpaper and screen-saver schemes that they generally don't bother installing today.

2) The new file system is interesting, but in a mixed environment -- one where new desktops are mixed with old, and in which the server environment may be Linux or some other non-Microsoft environment it may not work all that well. It also remains to be seen how user-defined lists will work in environments where people share files, and almost by definition, where they need to be able to share common, standard locations, ie folders.

3) Built in DRM schemes. Yawn. Most corporate networks have music downloads and streaming audio of all sorts blocked at the firewall. Who cares?

4) Built in search. Again Yawn. Who doesn't already have a Google searchbar. It's one of those plugins that I explicitly don't block people from downloading, and in some cases have gone to the trouble of just installing it on every machine.

5) Networking changes and support for IPv6 are something Microsoft should have done a while ago, but won't be all that important behind a corporate firewall.

6) Security improvements. Also long overdue and important to the home or mobile user, but virtually everything they're now building into the operating system is already built into every corporate firewall in existence. Within the firewall, some of these features may actually be disabled, as they'll most likely add complexity to maintenance and troubleshooting.

Nothing Requires it Yet

Historically, new operating systems have been adopted for two reasons. The first is that some kind of new software required the latest and greatest operating system to run on. Since 2000, that has changed substantially. While there's lots of software that won't run on platforms developed prior to Windows 2000, most current offerings are fairly compatibile with Windows 2000 and up. And there's nothing so new about Vista that is likely to change that. Microsoft will stop supporting good 'ol Win2K, but lots of companies will still use it for a while as it's stable and supports everything they need to do. Migration is likely to be slow, especially in companies that are already running Windows XP.

Another reason for and upgrade cycle has often been that the combination of the new OS with newer hardware allows corporate users to do things they couldn't before, and to be more productive.

This is a facet of computing that the web, and web-enabled applications has changed forever. With more and more corporate enterprise applications running in a browser or through some other thin client, there is less and less reason to upgrade the desktop capabilities. Internet technologies have made the desktop environment less and less relevant. Only the most extreme users of packages like Excel and Access will come close to taxing the capabilites of even three and four year old desktop systems. More typical Word and email users won't even come close.

Until some new "must have" application comes along that requires a faster desktop and a better OS, corporate IT folks will be mostly happy to continue with the stuff they have and that they know works.

Current Licensing Schemes Don't Force Change

Several years back Microsoft changed the way they did things. In the past, corporate customers paid to upgrade their old software with newer versions. You could skip a version or two and still get the "upgrade" price. Eventually, they'd cut off the upgrade program for whatever you were running and you were forced to upgrade or be faced with the cost of buying at the regular price at some point in the future. This tended to result in waves of upgrades, in which corporate buyers waited as long as they could, then purchased the upgrade rights. They didn't necessarily install the new version, but they purchased it. Microsoft's revenues tended to be "lumpy" -- long lulls followed by shorter and sometimes rather violent upswings.

That hasn't been the case for several years. Now, corporate customers pay annual license fees. Keep paying the fee and you always remain entitled to the latest version. Stop paying and you'll need to buy everything at cost whenever you want to make a change. Most corporations pay the license fees even if they don't plan to upgrade for several years. Some companies will continue using old, unsupported versions where it makes economic sense to do so. In my last full-time job we still had dozens of users on workstations running Windows 98 (not 98SE, plain vanilla '98) because a changeover would have required completely re-writing the in-house application that those people used almost exclusively. Given the choice of continuing to use the old thing or spend lots of money to upgrade to systems that would be employed in exactly the same way, we just stuck with the old. As far as I know, they still do, though a long-awaited CRM implementation may, finally do away with them.

As a result of this, Microsoft's revenue stream has become a lot more stable and more predictable. In fact, they told the world that stabilizing revenues was the primary reason for making this change around the turn of the century. Yet still, on Wall Street you hear talk about the power of the upcoming "upgrade cycle," by analysts who are apparently still not aware that Microsoft has done everything in its power to stamp out that cycle and has been pretty effective at doing so. There still is a cycle of upgrades within most businesses, but that cycle is no longer impacting Microsoft's revenues very much. You pay over time for the right to the latest version regardless of whether you choose to implement it now, next year or never.

Even Gartner, who are usually cheerleaders for any new technology, are suggesting that it won't be until 2008 that we see a strong move to upgrade current corporate systems. And that's going on the assumption that corporate IT folks make it a priority.

Corporate Users Aren't Everything

The consumer is increasingly important and has a greater impact on the purchases of PCs and operating systems than every in the past. But even here, the evidence of a strong upgrade cycle hasn't really been all that evident since the 1990s.

Certainly Windows 95 was worth upgrading to. And many users happily also upgraded their hardware to the new state of the art (then, the original Pentium) in order to jump onto it. Windows 98 was much quieter a release and few people upgraded from Win95, since it really didn't do all that much different. Win98SE was never even really available as an upgrade and few bought new hardware just to get the operating system that came with it. Windows ME was a complete abortion and I know many, many users who decided to upgrade because of perceived but nonexistent Year 2000 problems who regretted the day they inserted the WinME upgrade CD into their formerly reliable Win95 or Win98 machine.

Even WinXP -- arguably the best thing ever to happen to the consumer desktop -- hasn't been all that big a deal as an upgrade item, this despite the fact that the price is right and the upgrade process -- unlike previous ones -- ran pretty flawlessly.

So I don't anticipate lots and lots of consumers running out to buy the latest operating system. It's just not all that big a deal to most of us, even more sophisticated techies like me. I'm sure I'll end up with a Vista machine in time, but just not sure how soon that time will be. Probably not until Microsoft puts out at least the first wave of patches and fixes, and most likely not until they put out the first complete Service Pack.

Certainly, there will be some who feel the need for a new machine and will use Vista as an excuse. Microsoft's timing for the release -- in the second half of 2006 -- will probably tend to spur back-to-school and Christmas spending to a somewhat higher level than would be seen otherwise, but I doubt it'll have that great an impact on Microsoft's increasingly predictable and steady operating system/office software franchise.

The Bottom Line

Microsoft set out at the turn of the century to change their business. They've moved away from the booms and busts of the old upgrade cycles to a sales and licensing regime that tends to promote slow, steady growth. And they've been fantastically successful at doing this. So successful that they've started paying greater dividends and moved their employees away from heavily stock-based compensation and into other schemes.

Wall Street needs to take a clue. The upgrade cycle as we have known it for the a couple of decades is dead. Microsoft is a good company and may even be a good stock (though it was certainly a much better buy two months ago). But those looking for the kind of upswing in earnings that we saw with new OS releases in the 90s are likely to be very disappointed.