IBM Goes Deeply Predictive, Announces Acquisition of SPSS

James G. Kobielus By James Kobielus

IBM dropped a big bombshell at the start of any already action-packed day for the analyst community. At this moment, I’m sitting, along with several dozen of my peers from Forrester and other firms, at the IBM Smart Analytics System launch event in Hawthorne NY. I’ll blog on IBM’s other announcements in a separate items.

The bombshell was IBM’s announcement that it’s acquiring SPSS, a long-established, leading provider of predictive analytics (PA), data mining (DM), statistical analysis, and text mining tools. The acquisition, subject to the usual shareholder approvals and regulatory reviews, is expected to close later this year. But, even in advance of that near-certain consummation, IBM’s bold move has already sent shockwaves throughout the analytics market.


Most important, IBM has acquired the second largest vendor of PA/DM solutions, dwarfed only by privately held SAS Institute. In this segment, IBM’s proposed acquisition is having the same impact that its Cognos buy had on the business intelligence (BI) market two years ago. In many discussions with Forrester customers, SPSS is often mentioned as a key solution provider for predictive modeling and statistical analysis against structured, semi-structured, and unstructured content.


For IBM’s competitive standing in the data management market, this acquisition represents one of the last missing pieces of its Information On Demand (IOD) portfolio. By acquiring SPSS, IBM has acquired a substantial PA/DM brand with a very loyal set of longtime customers who have build their customer churn, supply chain optimization, and other predictive models on its best-of-breed platform. SPSS recently underlined its feature-comprehensive value proposition through re-branding around the “Predictive Analytics Software” (PASW) family name. But longtime customers didn’t need to be reminded, of course, that what used to be known as “Clementine” defines a functional high-bar in this solution segment.


IBM would probably be the first to admit that it took its focus off the PA/DM market over the past several years as it build out the BI, data warehousing (DW), and other pieces of its IOD portfolio. IBM had never really exited the PA/DM market, but casual observers might have thought otherwise. However, the vendor three years ago chose to de-emphasize its Intelligent Miner tools--which support mining of structured data--as stand-alone offerings. It essentially buried these solutions, moving them into its InfoSphere Balanced Warehouse family, where they are now offered as features of its Enterprise Edition DW software, rather than as stand-alone tools that would be enhanced and evolved independently.


IBM and SPSS’s respective customer bases should rest assured that overlaps among their respective product portfolios are not extensive. Once the acquisition closes, IBM is almost certain to build out its SPSS brand and, over the coming 1-2 years, phase out the Intelligent Miner technology within its InfoSphere portfolio. One tricky issue is which text analytics solution family--SPSS’ or IBM’s OmniFind solutions--will prevail as the parent converges these offerings in its IOD portfolio. Another issue is how IBM will integrate the SPSS offerings into its still-evolving in-database analytics roadmap for InfoSphere Balanced Warehouse. Hopefully, IBM will maintain and extend SPSS’ already extensive in-database analytics integration with a broad range of vendor DWs, including such Big Blue rivals as Oracle, Microsoft, Sybase, and Teradata.


Who loses from IBM’s acquisition of SPSS? Fundamentally, one can’t help think that SAP missed the boat by not seizing the opportunity to acquire partner SPSS, whose Clementine technology it OEMs, has integrated with its BI technology, and sells as SAP BusinessObjects Predictive Workbench. PA/DM is an increasingly key component of a full-fledged BI solution stack. However, the remaining field of vendors with stand-alone, horizontally applicable PA/DM vendors consists primarily of vendors who are large but proudly and stubbornly independent (especially, SAS Institute); high-quality but much less widely adopted (e.g., KXEN, ThinkAnalytics); or specialized on customer, financial, scientific, or other specialized analytics (e.g., Unica, Fair Isaac, Accelrys).


Some IBM rivals in the BI space already have strong PA/DM tools, most notably Oracle (Oracle Data Miner) and TIBCO/Spotfire (the Insightful tools). Among BI vendors, Microsoft, MicroStrategy, and Information Builders have PA/DM capabilities, but they are not to a SAS or SPSS level of sophistication. These and other BI vendors should also be scouting for strategic acquisitions.


What do you think? Will IBM’s acquisition of SPSS lead to further merger and acquisition activity in this space as other leading BI players strengthen their PA/DM solutions?

Comments

re: IBM Goes Deeply Predictive, Announces Acquisition of SPSS

Jim, as always - great post! One point I would add is that there are still plenty of missing pieces in IBM's IOD (and all other leading BI vendors) portfolio. In memory analytics, post information discovery, process analytics are just some of the examples. I do not see this acquistion to be the last one by any means. Cheers!

re: IBM Goes Deeply Predictive, Announces Acquisition of SPSS

Boris:For sure. No analytics platform/tool suite can ever be truly "feature complete," because the industry functional "high bar" continues to evolve. Innovations often originate with fast-moving startups and then, later, are absorbed by the larger, diversified vendors. As you say, expect plenty of new M&A, partnering, and OEM activity going forward.Jim

re: IBM Goes Deeply Predictive, Announces Acquisition of SPSS

The consolidation in the analytics space will continue not just for the reasons Boris cited.The implementation projects for these BI and Predictive Analytics projects generate significant revenues for large services company like IBM so they have incentive to keep the M&A going.What I wonder though is how this portend for innovative startups in this area. If the major players are bought up, does it provide space for smaller players to grow or does it choke off the go-to-market opportunities as buyers choose an all-IBM or all-Oracle solution?

re: IBM Goes Deeply Predictive, Announces Acquisition of SPSS

Of course, IBM's incentive to keep doing strategic acquisitions is competitive differentiation. Another incentive is the need to serve customers more completely--and expand their licensing and services revenues--through functionally comprehensive and integrated analytics solutions. In other words, by being the proverbial "one-stop shop" for BI, data warehousing, predictive analytics, data mining, data integration/quality, master data management, text analytics, and all other components of the analytics universe.But let's not exaggerate the importance of a "one-stop shop" for analytics platforms and tools. Enterprise customers have made significant investments in best-of-breed solutions--in all of these categories--from diverse vendors. What users often want--more than a one-stop for solutions and support--is assistance integrating diverse vendors' offerings more seamlessly. Indeed, users are often wary of relying a single vendor--no matter how blue-chip--for such a wide range of critical apps and infrastructure. The "best-of-breed" multi-sourcing strategy will continue to prevail among many enterprises, though they may continue to consolidate their vendor/platforms in each category (e.g., migrate from 4 BI platforms to one).The best-of-breed multi-sourcing approach creates an ongoing opportunity for the innovative pure-play and startup to sell their wares into big accounts that are dominated by "one-stop shops" such as IBM, Oracle, SAP, and Microsoft. Also, the fact that no one vendor is as yet a single source of best-of-breed offerings in every analytics category contributes to these opportunities.Nevertheless, IBM's acquisition of SPSS fills one of the remaining gaps in its otherwise comprehensive analytics portfolio. Till now, IBM lacked a best-of-breed data mining platform/tool family. When the deal closes, we await IBM/SPSS' roadmap that describes how the acquired solutions will be brought more fully into the parent's Information on Demand (IOD) portfolio alongside InfoSphere, Cognos, and other offerings.