SPSS Is Not the Story; IBM's Vision for Analytics Is

The media and my fellow analysts have been breathlessly touting IBM's acquisition of SPSS ($1.2B) as some sort groundbreaking to a new era of analytics. I don't see it that way.

IBM has been collecting BI and analytics pieces for a few years now, beginning way back in pre-history with the acquisition of Ascential. This was eclipsed when they acquired Cognos ($5B), the largest BI vendor at the time. There have been quite a few other smaller ones, such as AlphaBlox, Applix, Exeros and others. And let's not forget ILOG a year ago. You can't put predictive analytics into motion without a Business Rules Management System to drive the process. If anyone thinks this acquisition of SPSS marks IBM's serious entry into analytics, they've been sleeping.It should be clear to everyone that IBM's intention is not to sell software, but rather, to sell complete solutions to business. It is obviously fashionable to talk about predictive analytics these days, and now IBM has the bonafides to do that, but no one else has the vast resources to attack a worldwide market with hardware, software and business consulting. The PriceWaterhouseCoopers acquisition provides IBM with not only an army of techies and implementers, but also with a vast repository to business/strategy consultants. I can say from personal experience that implementing analytics is only part of the solution. Unless you have the knowhow (and carry the portfolio) to operate as a change agent, your success will be attenuated. This is the real genius behind IBM's effort.

My takeaway from the briefing at the Hawthorne Labs this week is that IBM is offering an alternative to the data warehouse appliance -- a bundled hardware/software solution with the goal of shortening the "time to value." But where most appliance solutions rely heavily on the commodity and open source message, IBM's offering of the pSeries and DB2 is a true-blue stack. But more than an appliance, the thrust of the message is solutions delivered by their service organization.

I'm encouraged by this. The endpoint of BI should be decisions, not dashboards, but it seems to have taken an inordinate amount of time to get to this place. I've always felt ambivalent about IBM's role in the wider BI market, but now they have my attention.

The announcement that brought us all to Hawthorne was no world-changer and most of us thought it wasn't really worth the trip. I guess I agree with that, expect that I was able to spend some time with colleagues and meet some new ones as well as finally meet some people face-to-face that I've "known" for a long time. And for me personally, having never had much contact with IBM, these new relationships are very worthwhile. So I'm glad I made the trip.

Now what REALLY got me excited was the Watson Project. "IBM has unveiled the details of its plans to build a computing system that can understand complex questions and answer with enough precision and speed to compete on America's favorite quiz show, Jeopardy!." I want you to think about that for a moment. Deep Blue beating Gary Kasparov? That was child's play compared to this. The advances in natural language technology to pull this off are fearsome. I wish I could describe some of the things I learned, but obviously this is being kept under wraps. How about the ability search a knowledgebase of unstructured data and score possible answers in the time to click the gizmo?

One little bit of irony. IBM may have been slow to throw its resources into BI/Analytics, but back in the day, as they say, IBM was the only game in town. They invented the Information Center (heck, they invented the "Information Warehouse"), and in my early career, we did our modeling in those tools like APL and ADRS. Something happened and IBM became the symbol of expensive, proprietary, IT-centric operations while we all looked elsewhere to solve our problems. Now IBM is back. I'm OK with that.The media and my fellow analysts have been breathlessly touting IBM's acquisition of SPSS ($1.2B) as some sort groundbreaking to a new era of analytics. I don't see it that way... If anyone thinks the acquisition of SPSS marks IBM's serious entry into analytics, they've been sleeping...