The blogsphere (blogland? blogosity?) thrives on drama. It drive traffic, boosts egos and sells advertising.

One common source of drama is Vista in the enterprise. I am not going to get into the drama itself but instead here is a short note on what drives enterprise adoption of software and systems.

There are three things which CIOs want from software:

1. It should work
2. It should be supportable
3. It should be affordable.

Who could argue with such tenets? But when you unpick them you begin to understand why the world is the way it is and why, alas for drama, logic tends to be the main driver.

It should work

Even in my organization we have a lot of enterprise systems which we need to work. Some of them are off the shelf applications, some are bespoke internal applications and some are external systems which we need to use for line of business or cross-cutting activities.

If I change my desktop operating system then I need to retest all of those enterprise systems to make sure that they still work. Even in my organization where we have long since moved all bespoke applications to the web we still need to ensure that they work with the combination of browser, OS and security controls we have to put in place.

Let’s assume that I have 50 enterprise systems, each to be tested. They will vary in complexity and scope but you would do well to start with an estimate of 50 person weeks to test those systems.

Of course you could do some in parallel and of course you would look to drive down the time needed but remember the basic rule of management – “hope for the best, but plan for the worst”.

And don’t forget that many organizations have enterprise systems that are infrequently used. How often do you run year end financials?

It should be supportable

When someone has a problem they phone the Helpdesk. Fine, who does the Helpdesk phone?

If I install version n.0 of something then is there sufficient body of support for it or is it like the old story of the teacher who said that the key to success was to be 3 pages ahead of the class? Telling the business that the supplier is fascinated by the problem they are having with the profit forecast spreadsheet is not support, resolving the problem is.

So there will always be a lag between release and adoption. The length of the lag then depends on appetite for risk and how important that system is to your business. We adopt security patches very quickly, new desktop wallpaper not so much.

And then one of the ways we try and reduce the support overhead is to standardize on a standard image so every desktop, every laptop is the same as each other. But we have to build, test and then install the image. Fine, the supplier will do the latter but it still relies upon us to be confident that it will work.

So factor in the time to build, test and distribute the image alongside the time needed to test enterprise systems and time marches on.

It should be affordable

Moving to a new desktop costs money and uses resources. Even if the update was free there are still hard and opportunity costs around doing the upgrade.

If I am moving to a new desktop then I am not doing something else. What matters most to the organization? How do the costs of supporting the old world compare to moving to the new world?

So, despite our love of drama the reason for slow enterprise adoption of systems is usually just the result of cold logic.

You could replace Vista in the example above with Linux or OSX and the logic remains.

And as a final piece of context, I am writing this entry on a Vista laptop before posting it to the site which was set up in OSX and which I usually read on Linux.

I am sure that someone will mention enterprise agility and organic implementation at some point and I will return to those later.