I have a theory about deploying Web 2.0 applications within the enterprise. I’ll call it a theory simply because I can only prove half of it. Suppose we have the time, money, and resources to put up five implementations at once, each with a different software solution. Let's say we are going to deploy an enterprise wiki and we pick Social Text, Confluence, PBWiki, and two others. The software is the only variable in the mix which means that the amount of training and education are the same. The level of internal marketing and branding are equal. We also build the same self-service environment, automate the business processes, and provide the same client-support. Finally, the level of integration services provided by an experienced staff of professionals. After 12 months, I am going to say the difference in the amount of content captured within the environment and the amount of usage will be less than a 5% difference. That is to say that the number of employees, the number of pages, the number of views, and the associated attached documents will be less than 5% different.

Of course, I can't prove that since no one is going to allow me to do this, nor could I get an equal level of effort on my other variables. Now, let's turn it around. Suppose, I have four implementations where the product remains the same but we wildly alter the delivery level of our variables. What is the end result of this? You should be able to guess this one; there may be a 95% difference in the level of integration and usage. I can prove this one based on the four companies that we have integrated. The problem is that while few people can actually argue with this logic, they will still spend 90% of the effort on the software selection, infrastructure, and the many other components of failure. That is, the things we do so we won't fail. Yet, we do very little toward the components of success which are the real determinates of Web 2.0 integration.