John Wolpert described his activities at IBM. I personally find what he is doing to be remarkable, and I marvel that more companies don't do the same.

Here's why: when it comes to R&D, companies think nothing of designing various experiments to test out various possible technical combinations of features or components. Companies have become quite savvy in designing these tests, in monitoring their results, and in making important decisions based on the results of those tests. Companies even keep track of their portfolio of development projects, and have a sense of how the collection of possible future products is faring within the company. This gives them early warnings about a lack of new products down the road, in time for them to take corrective action.

However, those same companies have no similar set of practices outside of the technical realm, when it comes to the business decisions that they must make. There, the dominant logic is to proceed according to the business model that the company is currently operating. In many companies, this approach is called "feeding the beast", or "fueling the business". What it means in practice is, "if it doesn't fit our business, either change it so it does, or drop it".

The alternative to this myopia is to test alternative business models that might make better, more valuable use of a technology project. Yet this is what most companies don't know how to do, and have no process for doing. As a result, they have little or no sense of how they might grow beyond their current business. There is no visibility - or early warning - of a business reaching the end of its growth. And think of how many innovative, high technology companies there are these days that are desperate for growth!

Enter John Wolpert, and IBM's Extreme Blue. While I have not visited these guys in person yet, I have admired them from over the web and through casual conversations. And what this unit is trying to do is to experiment with business models - alternative ways of configuring technologies to serve possible different markets - the way that IBM experiments with future technologies.

And look at who they employ to do this experimentation. These are not your father's IBM employees. Often, they are only summer IBM employees. They are usually young, probably inexperienced, probably very energetic, and haven't yet learned what won't work. Precisely the kind of people you'd like to have doing experiments: people who don't already think they know the answer.

So I applaud Extreme Blue, and what it is trying to do. I wonder what happens to those experiments, once they are run? What is the "after action review" of these experiments like? How do you report results? How can results change minds and behavior in the Big Businesses within IBM? Notwithstanding, I commend this kind of experimentation to those of you who sense that there must be other ways to do business beyond your current model.