• To October 2003
• Since October 2003
• Office Trends
• Comment

Microsoft Office Version Survey

[Updated: 27 May 2004]

Why a Survey?

I make my living from Microsoft Office as an Office Applications Trainer and as a VBA Developer. When I first started using Microsoft Office in 1993 the current version was Office 4.3 and we were using Windows 3.11. When Windows 95 appeared Microsoft released Office 95. Then came Office 97, Office 2000, Office 2002 (branded "Office XP") and most recently Office 2003 (branded "Office System"). Microsoft has catered for Mac users too, with versions 4.2, 98, 2000, v.X, and 2004.

I haven't had any requests for Office 4.3 recently (although I still have the stack of 31 floppy disks on my office shelf in case anyone asks!) and Office 95 never seemed to be much of a success. But versions 97, 2000, 2002 and 2003 are all "current" and whilst Microsoft might be considering withdrawing their support for the earlier versions, professionals like me really need to be able to work with any and all of them.

For this reason, I wanted to know how much time and effort to devote to matters of compatibility between versions. As I write this (in January 2004) I am working on my "main" PC using Office 2002 with Windows XP. I have another two PCs on my desk, one running Office 2000 and the other running Office 97 (both on Windows 98). The Office 2003 disks are still in their box waiting for me to decide what to do with them!

I know from my clients, correspondents to my web site, and various discussion groups that people are using all four of the more recent Windows versions. Microsoft estimate that there are around 400 million Microsoft Office users worldwide, but I have not been able to find any published information on the comparative popularity of the different versions, or whether (as one might expect) the use of the older versions is in decline.

So, in July 2003 I began to collect my own data using a voluntary survey of visitors to my web site. Visitors were asked simply to state which version of Microsoft Office they were using.

^ top

Survey Results:
July 2003 to October 2003

By October 2003 over 500 "votes" had been received. On 21 October 2003 Microsoft released Office 2003 (in fact the first vote for this version was registered on 20 October). I decided to break the survey at this point on the basis that some users would be upgrading to the new version and earlier votes would become invalid. Here are the results to that date:

^ top

Survey Results:
Current Survey since October 2003

[Updated: 27 May 2004]
This survey commenced on 20 October 2003 to coincide with the release of Office 2003. The data contains only votes registered after that date:

^ top

Trends in Microsoft Office Version Use

[Updated: 27 May 2004]
The charts shown above give a "snapshot" of the distribution of version users. The following chart gives an indication of the trends that have taken place during the period of the survey. Each point on the chart shows the percentage of users voting for each version during that month.

^ top

Comment: What Does It All Mean?

The Results: Before Office 2003

When trying to make any serious conclusions about the data I have collected you should bear in mind my comments (below) on sample size and the people who supplied the information.

Before the release of Office 2003, the numbers of Office 2000 and Office 2002 users were more or less equal at just over 37% each. Office 97 was in a strong third place with nearly a quarter of the users (over 23%). This was testimony to the popularity of Office 97, with many people choosing to remain with it over 6 years after its original release. The strength of Office 2000 standing equal with Office 2002 suggests that many users upgrading from Office 97 have either moved to Office 2000 or waited and then moved to Office 2002, rather than upgrade as each version was released. Some people will be using Office for the first time and will not have upgraded from a previous version.

For systems managers in large businesses a version upgrade is no small matter. They have to consider not only the cost of upgrading their users but also weigh the benefits of upgrading against the requirements for training users in the new software. Nevertheless, as each new version of Office appears, it seems that there is a ready market for it.

As I expected, the numbers for the earlier versions are very small but this doesn't necessarily reflect accurately the "global" picture (although I think it probably does!). My web site is largely (but not entirely) devoted to VBA which did not become an integral part of Microsoft Office until version 97. Macro programming with Office Basic was possible in earlier versions but it was not as widely used as it is today. So users of earlier versions are less likely to be visiting the site and participating in the poll. I might argue that someone with an interest in VBA is also more likely to want to keep up to date with developments in the program and therefore more likely to upgrade (but I don't think I will!).

The number of Mac votes is small but significant. The Mac still dominates the design world and with the advent of the iMac has gained considerable popularity in the domestic market. Again it is necessary to consider who is likely to be visiting my web site. It is largely biased towards Access and Excel and, whilst I am sure that these programs are used by Mac users, I think that the survey under represents their numbers.

The Results: Since Office 2003

[Updated: 27 May 2004]
It is seven months since the release of Office 2003. Following a slow start, Office 2003 is climbing steadily and scores a significant proportion of the votes at 16%. Some of this number is likely to be "new" users of Office. Office 2000 and Office 2002 are almost equal at a little under 35%. Early results suggested that Office 2000 was declining faster than Office 2002 but the positions soon reversed and are now effectively equal. Office 97 is falling steadily, now standing at just under 13%, suggesting that an increasing number of Office 97 users are deciding to call it a day and upgrade.

Office Trends

[Updated: 27 May 2004]
The two polls described above each use data collected over a period of several months. To get a clearer idea of the trends in Office version use I plotted the data collected each month, showing each version as a percentage of that month's votes.

Initially, the results indicated that Office 2002 was declining in relation to Office 2000, but the programs now stand about equal. I would have expected that the percentage of users of the older version to decline as more people decided it was time to upgrade. Perhaps Office 2002 users are more likely to be the sort of people who like to keep up with the latest developments, and that the decline in Office 2002 results from upgrades to Office 2003.

After a slow start, the percentage of Office 2003 users increased steadily for a while but now the rate of increase seems to be trailing off. Office 97 displays a corresponding decline, which also seems to be levelling off. I had though that maybe Office 97 users had at last decided to throw in the towel, but this version clearly has a lot of devoted users.

All this tells me that I still need to consider users of all the established versions (97, 2000, 2002) and that this is likely to continue for some time. I still haven't been prompted to move to Office 2003 although the software sits in its box on my desk. From a user/developer's point of view I see no need to move from my usual Office 2002.

The Sample Size

[Updated: 27 May 2004]
I didn't consider it worth publishing the results of this survey until the poll (the number of people who had registered a vote) exceeded 500. As I write this the poll has exceeded 1600 votes and, in my opinion, means something. But bear in mind that Microsoft estimates there to be around 400 million Office users worldwide! I make no claims that this survey represents anything other than a cross section of my readers.

Who Voted?

I do not present this survey as a representative picture of the worldwide use of Microsoft Office, although I think it might well be - and since nobody else is offering one it's the best picture you have. But it's relevant to me. Those who voted are restricted to people who visited my web site, who happened to notice the voting form and who were interested enough to vote. This means that they aren't exactly your average Office user. They are more likely to be someone in the category of "power user" who uses Microsoft Office a lot and wants to learn more about it.

What About Vote Rigging?

Every vote is scrutinised (by me!) and there is a mechanism to spot multiple voting. It is reasonable for a person to vote for two different versions when, for example, they might use one version at work and a different one at home. Someone recently attempted to vote twelve times for the same version, and whilst it is possible that they were voting on behalf of their entire office, I decided that only one vote would count. And no, I'm not going to tell you how I knew it was the same person voting each time!

©2004 Martin Green. All rights reserved. martin@fontstuff.com