jump to navigation

Upgrade now or upgrade later, that is the question May 6, 2007

Posted by Matsu in Business, Information Technology, Management, Microsoft, Software, Technology, Uncategorized, Windows/Microsoft.

A good friend of mine has a favorite phrase, “Timing is everything.” While they may not be a technology professional, they certainly know the secret to upgrading systems — timing. Performing upgrades at the right time makes all the difference. So, when is the right time? When should you switch to new versions of applications or operating systems?

I am presently faced with that question where I work. Like the rest of the world, we must decide when it is best for our organization to upgrade to Microsoft Office 2007 and Windows Vista. Of course, they are two separate decisions as you can install Office 2007 on Windows XP and, as far as I know, you can also run Office 2003 on Windows Vista (though I don’t know why one would choose to do that).

While I talk about Office 2007 and Windows Vista upgrade in this post, these questions and decisions apply to all types of software, even open source software. The concerns and problems (or opportunities) are just as real with Apple’s Mac OS X and any flavor of Linux. So, you may want to continue reading even if you don’t use Microsoft Office or Windows.

To answer the upgrade question for your organization, you must consider both business and technology factors. But, beware! The one reason that should never be used when deciding if to upgrade is because it’s “new and shiny.” As in, “it’s new, it’s cool, so it’s got to be better.” Some technology professionals may fall prey to that siren call, but you shouldn’t. And, if your top management gives you that argument, be sure to give them a dose of reality by providing a list of pros and cons for the upgrade. That will let them know you are seriously considering it and looking at it from all of the angles.

First, let’s consider the business opportunities. Business reasons include performance improvements, which may mean higher productivity, and greater security, which may mean less risk (and in the case of viruses, it can mean less down time). Of course, there may be business reasons to upgrade because of new functions or features that provide an advantage over the current (older) version of the software. And, in today’s highly regulated business environment, it’s possible that the new version of software is necessary to be compliant with some new government regulation. Those are just a few reasons that an upgrade would be needed or advantageous.

There are also business reasons not to upgrade. For one, it may require the users to learn a new way of doing the same old things. What I’m talking about is a change in the software’s user interface. That is certainly something that applies when considering upgrading to Office 2007. The user interface is radically different from previous releases of Microsoft Office. Microsoft argues that while there is a learning curve to their new version of Office, once learned it is more intuitive and users will do things easier or faster. Hmmmm. It’s too early to say how true that may be. And, if you don’t have time for your users to relearn how to use a word processor or spreadsheet, then maybe you should wait for a time that would allow people to learn the new software without impacting (dramatically) their productivity. (There can be a whole blog post about training, so I won’t go into the many options and best practices here.)

Another reason not to upgrade is the cost. You not only must consider the upgrade license costs, but also any hardware costs (if better, newer, faster, hardware is required). That may not apply to Office 2007, but it certainly does apply to Microsoft’s new version of Windows (Vista). Probably the greatest reason that businesses are holding off on rolling our Windows Vista is the fact that newer hardware is a requirement.

According to this story by Cnet News, Dell and Intel are holding off on upgrading to Windows Vista. The reason for their decision is largely based on the stability and security concerns of a “1.0” release of software. Therefore, they are waiting for the first Vista update or service pack (SP1) before conducting a large rollout of Vista to employees. I think that it shows both a level of understanding of technology and the business impact to make that decision. And, I believe it’s the right decision for them.

While I have not highlighted technical reasons to upgrade or not to upgrade, they do exist. In my discussion of the business reasons they had foundations in technical reasons. After all, you can’t totally separate the technical from the business impact or potential opportunities. Just to cover all of my bases, you should consider the following technical issues when determining the best time to upgrade: license management (Vista is different and significantly increases administrative overhead just for taking care of licensing), impact on infrastructure (servers, network, directory services), and compatibility with existing systems (legacy applications, mission critical functions, etc.).

Of course, you should not make the decision to upgrade any application or operating system without thorough research. That means testing the applications on your systems (test systems, not live/production systems). After determining all of the upgrade implications then you must address any compatibility or performance issues that you discover. After all, you do not want to find out halfway through your deployment that some mission critical application or process won’t run. Only after a complete discovery of all problems should you consider the best time to upgrade your organization’s computers.

I have already stated that the timing must be determined by each organization because they all have different needs and issues. But, for my organization we have established some best practices when it comes to the timing of an upgrade. We have established a policy that we don’t install new application software on end-user’s computers until at least three months from the release (ship) date. And, for operating system upgrades, we extend that timeline to approximately six months. Again, this is a guideline or ‘rule of thumb’ but we tend to stick fairly close to it. By the time an application is three months old or an operating system is six months old the marketplace has found and possibly fixed any problems like compatibility issues, security flaws, and performance shortfalls.

One final piece you should consider relates to technical support. Waiting three or six months gives the technology support staff sufficient time to ‘play’ with the technology and become more comfortable using the new software. More experience and comfort with the new software translates into better support.

So, when do you plan to upgrade all of the computers in your organization to Office 2007 or Windows Vista?



1. Bruce Byfield - May 6, 2007

I can’t help noticing that you reject the idea of not upgrading.

Why is that? Is there anything that the new software versions do that your old ones don’t?

2. Matsu - May 6, 2007

Bruce, you raise an interesting point.

It’s not that I reject the idea of not upgrading since that is always an option. But, the reality is that eventually everyone will upgrade. Ask someone who is running their entire business on Windows 95 or Windows 98 how difficult it is to be so far behind in technology. And, organizations that run Windows NT on all of their Intel based computers know that Microsoft has stopped releasing security patches, so they are at risk.

While it may be a perfectly viable option to put off upgrades for a couple of years, eventually an organization with a large group of users that rely on computers will have to upgrade their software (no matter which applications or operating system they are running).

Therefore, I see ‘not upgrading’ as a short-term option (one to two years) and long term (two or three years) organizations will have to move their applications and operating system forward.

Of course, everyone must answer your follow-up questions when deciding to upgrade or not. I believe that ultimately it’s not a question of “if” but “when.”

3. Moose - May 7, 2007

I would agree that in your organization “no upgrade” for client software is certainly not a long term option. Unfortunately, this is probably also true for non-client software as well, though the time-lines are much easier extended.

Upgrades should be a part of a formal project management plan. You hit on some of it above. I am not a huge “formal project management” proponent for most things IT – as the time it takes do this is not worth the efforts for a good number of IT projects. A less formal PM plan works better.

About a week ago I got Project Management Institute’s PMP book “Project Management Professional Examination Specification” (http://www.pmibookstore.org/PMIBookStore/productDetails.aspx?itemID=750&varID=1) which gives a great outline of project management from a task outline perspective.

I would recommend it if you do not already have a good project management “shell” document.

Anyway, you are in a difficult position. I remember when I was doing work at Vandy and got a good look at what they had to support there – multiple OS’s, multiple applications, etc. It is an amazing and difficult job. Just the training of internal staff and the documentation associated therein is extremely time consuming and difficult.

I can’t imagine where you have an option of not upgrading things like the client operating system, the client production tools (Office, in particular), etc. In addition, due to security concerns, bugs, etc.; I can’t imagine not upgrading Server side stuff either – which is completely different with its own set of issues and effects on the community.

Yours’ is the most difficult environment on the planet. Your users range in intelligence from “normal” to “genius”, your users are in an environment where they are being encouraged to grow intellectually, thus things like expirimentation and research are the mindset, your users have more money to spend on non-essentials than any other group of people, your users have lots of time on their hands that they must fill; and , finally, they are in an environment (most for the first time) where they are making decisions on thier own without the direct scrutiny of authoritative figures (parents, grandparents, etc.) and are, thus, learning where the boundaries are and are not for themselves.

Oy ve! I believe the phrase, “damned if you do, damned if you don’t!” applies here.

4. Bruce Byfield - May 7, 2007

Whether you need to upgrade really depends on what your business is. Even today, there are numerous small businesses running on DOS and Windows 3.1.

I admit that there are tremendous pressures to upgrade in some fields, especially technical ones. However, much of this pressure is a matter of prestige rather than necessity. The average office user’s needs have been filled quite adequately for over a decade, and, in many cases, it is only the wish to appear current that drives updating.

Another point, of course, is that if you are thinking of upgrading anyway, you might want to consider other alternatives, such as GNU/Linux, which is now perfectly suitable for typical office use, and tends to use incremental upgrades, which are a lot less onerous for those who have to manage them.

5. Matsu - May 7, 2007

Bruce, I agree with your argument for small businesses. In fact, to support your argument I offer you the automobile. If a car is seen simply as a means of transportation and if when the car is new it takes you from point A to point B, then it’s a good car. Over time, say about 10 years, it can still do a good job of getting you from point A to point B, provided you maintained it well. Most of the time, people who purchase new cars and drive them for a few years then replace it with another new car are not doing so to get ‘better transportation’ but instead they do it for other reasons, such as appearances or to feel better (vanity?). Generally speaking, a 10 year old car can do just as well as a new car when it comes to driving to work or the grocery store.

In the same way, a 10 year old operating system can still do what it used to do when it was new and some small businesses can get away with running old applications and very old operating systems. But, larger enterprises simply can’t do that. Their ERP systems may require newer operating systems or their manufacturing software may require newer hardware (and newer operating systems). There are so many more external forces pushing large enterprises than with small businesses.

So, I agree with you but argue that it only works when a business is operating on a smaller scale and can be somewhat isolated from those external forces that affect the large enterprises.

6. Karl Craig-West - May 19, 2007

Great blog article. Couldn’t agree more.
I used to be an IT Manager and wish I’d had access to such common idea and information sharing such as this.
Keep up the good work.

7. Matsu - May 19, 2007

Thanks, Karl. I hope you visit in the future and gain something new from my postings.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: