Upgrading for Upgrade's Sake

When is an upgrade not an upgrade? All too frequently, moving to the latest version of a software package can bring as many problems as benefits.
Posted August 29, 2001

Drew Bird

Quite why Microsoft considers the release of the next version of Windows worthy of a Y2K style countdown is unclear. What is clear is that once again we are faced with same question that has surfaced many times: What, if anything, is to be gained by upgrading to a new version?

Most software companies, not just Microsoft, release new versions and products with what seems frightening regularity. Since there is no policy that governs what is termed an upgrade, the number of new functions built into an "improved" product is at the discretion of the manufacturer, as is use of the term "new version."

Whatever the changes, the release of a new version seems to bring with it a mentality that upgrades are essential. But are they always needed, or are we often being lured into an "upgrade for upgrade's sake" mentality?

One IT manager, who requested anonymity, provides an interesting perspective on the subject: "Each time a new version of a product comes out, our IT director is keen to get started on it as soon as we can. He feels that it presents the right image to customers and shareholders if the IT is cutting edge."

But I am sure that customers would rather see lower prices, and the shareholders more profitability, than a yearly rollout that costs a couple of hundred thousand bucks.

Only a small portion of that cash relates to actual software purchases, which tends to be the first question in the 'shall we, shan't we' upgrade process. The cost of buying software for upgrades pales in significance compared to costs associated with product deployment. Expensive technical staff spend weeks or even months preparing for and carrying out the deployment. Hardware often needs to be upgraded, and in some cases end users need training. The costs can be staggering.

On the subject of end users, it's worth mentioning that the upgrade mentality is not limited to the server room. Just recently a colleague suggested I upgrade to Office XP. Why? "Because it's new" was the answer. Needless to say, I didn't upgrade. I probably use about 10 percent of the features of my word processor, and I barely use the other tools in my office suite at all. You could give me a 486 with Wordstar on it, and I'd still get by.

There are, of course, instances when upgrades are justified. Back in the mid-1990s, users of NetWare 3.x realized the phenomenal benefits of upgrading to NetWare 4.x. The same could be said to a lesser extent for those moving from Windows NT4.0 to Windows 2000. In these cases the argument for upgrading was strong.

That said, both NetWare 3 and Windows NT4 are robust operating systems that, in some environments, offer everything that's needed by an organization. If there were a good reason for not using these older systems, it would be the lack of ongoing technical support and the availability of hardware drivers. Outside of this, there are few reasons why an organization that needs basic file and print services on a small number of servers would need anything more.

So frequent are new product releases, and so time consuming the rollout process, that many businesses find themselves spending more time upgrading products than actually using them. Again we return to our anonymous IT manager.

"The worst thing about the perpetual upgrade cycle mentality is that you don't get a chance to stand back and really find out what the existing system will do," the manager said. "We use the same services of the system we just installed as we did on the previous one, even though it can do so much more. That makes the fact that our bonuses were cut due to the rollout going over budget hard to swallow."

The actual determination for an upgrade should boil down to one simple thing: If an upgrade provides functions that enhance the business process and brings benefits that will outweigh the expense of buying and installing it, it's valid. Anything less, and the answer has to be no. Then you can spend all the time you've saved watching the Windows XP countdown clock.

Drew Bird (MCT, MCNI) is a freelance instructor and technical writer. He has been working in the IT industry for 12 years and lives in Kelowna, British Columbia. This article first appeared in CrossNodes, an internet.com site.

0 Comments (click to add your comment)
Comment and Contribute


(Maximum characters: 1200). You have characters left.