Showing posts with label predictive. Show all posts
Showing posts with label predictive. Show all posts

September 18, 2009

Predictive Analytics, Cloud Computing, and IBM

Recently I've mused at the planned acquisition of SPSS by IBM and the possibility of bringing predictive analytics in a completely new and radical way to the business audience similar to what Microsoft did with Farecast in Bing. In a recent article in Forbes.com, Erich Clementi, IBM GM for cloud computing revealed that
"In the next six months, IBM plans to extend its cloud computing offerings to include a range of business analytics tools"
Business analytics tools on top of cloud computing is certainly interesting, but not totally new. A few small and big BI software providers are already hard at work developing their SaaS offerings. Note: See an example of Cloud BI with Good Data - a Marc Andreessen's funded company.

The fact that IBM is serious about cloud-computing and already publicly mentioning business analytics as one of the cloud services is worth taking notice. The interesting part is what could happen in 18-24 months from now, when SPSS integration is essentially completed and new development projects well under way. Could we see Cloud Predictive?

It is certainly a far more creative possibility than just selling SPSS enterprise software licenses using a direct or indirect channel. No doubt IBM has made a number of interesting moves and announcements in the last few months.

Permalink

August 21, 2009

Can IBM Hit a Home Run with SPSS?

Bing, Microsoft's new search engine offers a very interesting travel search capability. At first sight, Bing travel search engine is not so different of what you would find on Sidestep, Kayak or Mobissimo. You type your from-to flight information and get a selection of flights to choose from. What sets Bing travel search engine apart is his predictive capability. Bing predicts if the price will go down, up or hold steady. It even provides a confidence interval (expressed as a %).



Microsoft has acquired this capability from Farecast about a year ago for a cool 115 millions. Beside the steep price and the debate around its valuation (read the The Value of Statistics by Ian Ayres and Expected Value of Information), travel is nevertheless a clever application for predictive. It will be interesting to see what's next for the Farecast folks at Microsoft.

It makes me wonder if IBM could do a similar trick and bring SPSS technology to the business audience the same way Microsoft did with Farecast for the consumers / travelers. SPSS has been the big gorilla in the predictive space and it's not too difficult imagining Big Blue being reasonably successful with this acquisition.

As one of the dominant infrastructure players on the planet, IBM has likely heard the call for help coming from customers with tons of corporate data having difficulties figuring how to leverage it for their decision making. IBM can and probably will attach SPSS to its BI portfolio, but whether or not they can integrate predictive analytics to other technology assets remains to be seen.

As Timo Elliot, one of my colleagues has pointed it out in his blog, it would probably make more sense to embed predictive capabilities into business applications. Trouble is IBM is not really a provider of business applications. Unless there's a major strategic shift, the elusive predictive 'home run' will have to come from somewhere else...

Permalink