Imagine, for instance, taking a typical web retailer as an example, being able to analyze clickstream data combined with other data sources, like call center data and warehouse inventory data - all in real time!
I am also assuming that this solution is not an ASP solution - you can capture, store and crunch data on your own machines. You can also certainly, explore other opportunities to store your data elsewhere - like in Amazon’s Ec2 cloud.  But the key point here is that it completely negates the need for a CPM type deal offered by current Web Analytic vendors.

A similar, but completely free version of this software, known as Project R has also been making waves lately. It seems to have the capability to do some fancy stuff with data. But be prepared to roll up your sleeves to do some coding. This is because I assume that both I.B.M’s software and Project R do not have have all the rules to define standard Web metrics built in by default. You will have to define and code them yourself.

But that may not be such a bad thing. You get to completely define any metric you want, in a way that is relevant to your line of business. You get to cut and slice data anyway you want. In the end, an organization will have to weigh the cost and benefits of going this route versus getting a polished off the shelf solution offered by the likes of Omniture.

Web Analytic vendors are also at a crossroads at this point.  The trends indicate that the future looks something like this - where the requirement is to combine and massage large data sets, beyond just web clickstream data, to provide a global view/insight into a company’s operations. So the question for them is - do they have to evolve into something like what I.B.M is offering (and SAS is also actively in this space) or continue to provide enhancements to the way just web clickstream data is analyzed. It seems obvious to me the former option is the way to go, and in that case, they are behind the curve.