Comments on: IBM Acquires SPSS – A Big Deal, or Not?

By: Andreas

Andreas — Sat, 03 Oct 2009 12:59:34 +0000

I believe in data-mining and predictive analysis. A good example is finding the products that sell well with each other/cross-selling, which can be achieved with data-mining. SAP to my knowledge has nothing of their own to offer at the moment for data-mining/predictive analysis. They have an OEM partnership with SPSS though.
I believe that data-mining is part of BI and including that functionality in one’s BI product suite will distinguish oneself from the competitors.

By the way the same as does real dashboarding, as part of visualization information, and not those often poor, not to say lousy dashboard examples one sees so often.

By: Tim

Tim — Wed, 30 Sep 2009 05:19:30 +0000

The only way that IBM can realize the value of the SPSS purchase is to open source the SPSS software (such as their PASW Statistics and PASW Modeler) or release it free to the community,just like their Eclipse IDE.

By: Pierre Leroux

Pierre Leroux — Sat, 19 Sep 2009 06:24:19 +0000

Hi Timo,
Thanks for the post and some good background research. I’ve recently mused about the possibility for IBM of making a bold move with predictive the same way Microsoft did with Farecast (bing.com/travel) not too long ago.

At the time, I too remained unconvinced that the SPSS acquisition was a seminal moment, but a recent article in Forbes.com caught my attention. Erich Clementi, IBM GM for cloud computing revealed that “IBM plans to extend its cloud computing offerings to include a range of business analytics tools”. Hardly groundbreaking you might say, but once IBM has digested and integrated SPSS, we could perhaps see something interesting: Cloud Predictive…
http://launchpad604.blogspot.com/2009/09/predictive-analytics-cloud-computing.html

By: Michael Zeller

Michael Zeller — Thu, 13 Aug 2009 17:49:54 +0000

Great post, Timo. I think IBM’s move highlights the importance of Predictive Analytics, but as you pointed out, often the emphasis is on data mining and building predictive models rather than how to embed intelligent decisions into the business process. That is where the real value of Predictive Analytics will come to fruition.

In addition, open standards like the Predictive Model Markup Language (PMML) now provide interoperability among many vendors. For a summary, please the recent panel discussion at KDD 2009 which was focused on production deployment and integration of predictive analytics in operational systems.

KDD 2009 Panel Report: Open Standards and Cloud Computing
http://smartdatacollective.com/Home/20029

By: Twitted by nenshad

Twitted by nenshad — Thu, 13 Aug 2009 01:57:32 +0000

[...] This post was Twitted by nenshad [...]

By: Merv Adrian

Merv Adrian — Wed, 12 Aug 2009 15:34:25 +0000

Timo, nice work as always, and thanks for the mention. I agree that it will continue to be challenging to drive adoption of advanced analytics, data mining and predictive analytics to wider audiences; history does indeed show that many have had your experience of bringing capable products to market and watching them languish. I believe it is all about context – bathing the user in an environment that reflects an understanding of the business issues and industry environment, together with guided tools that advise on the techniques and tools to select within that context will go a long way to breaking this perpetual logjam.

Everyone wants to get there, and there is little doubt that vendors with deep business process understanding, industry knowledge, suitable technology componentry (their own, or perhaps partners’), and available services (again, partners may provide this) will take the lead. The level of richness we are talking about, at a sufficiently granular level (meaning microverticals and fine-grained business processes) will make market success for smaller firms difficult, and the reward small. The individual products will have relatively small target markets so a vendor with a portfolio strategy, where much of the underlying technology is reused from one offering to another, will have the best chance of achieving financial viability.So acquisitions by the kinds of vendors you suggest – of the kinds of vendors you suggest – are clearly the next phase.

I made the point, also referencing Mary, about IBM’s relative lack of succesful rampup in its first year in my post. And that’s why my language was rather specific: IBM, I said, “has *signalled that it believes* itself ready to redraw the BI map.” [emphasis added] Most of the rest of the piece is about the same industry drama you’re predicting, and I share your belief that it will not be easy to turn decades of resistance to advanced techniques to broad success overnight. But we *are* having the conversation again, and perhaps we’ve learned something from those decades. If we have, then predictive is indeed what’s next – it’s the right objective, and maybe we can get everyone ready for it.