Analysis: IBM’s $1.2 Billion Acquisition of SPSS

Big Blue’s acquisition of SPSS could herald a return to prominence for statistics gurus, model architects, and other power users. It’s about time, industry watchers say.

IBM Corp. yesterday agreed to acquire analytic stalwart SPSS Inc. in an all-cash deal valued at $1.2 billion. That’s almost 50 percent more than SPSS’s prior-day closing price -- and nearly four times its 2008 revenues.

The upshot, analysts say, is that IBM’s bid for SPSS amounts to still another demonstration -- if any were needed -- of the increasing importance of analytics.

Call it the analytic-i-zation of the enterprise.

There’s an additional wrinkle here, too: analytics isn’t a turnkey -- or entirely technology-based -- proposition, industry watchers stress. It’s in this respect that Big Blue’s move (considered alongside similar analytic-centric initiatives from Oracle Corp., SAP AG, SAS Institute Inc., and other players) underscores the importance of human know-how. For this reason, some experts suggest, the analytic-i-zation of the enterprise could herald a return to prominence for statistics gurus, model architects, and other mathematically-gifted power users.

IBM Out in Front

Big Blue itself has been an eager proponent of enterprise analytics: earlier this year, for example, it unveiled its Business Analytic Optimization (BAO) services, a dedicated analytic offering that prescribes -- in addition to (preferably) ample doses of IBM business intelligence, data warehousing, and integration middleware technologies -- the top-to-bottom reorganization of a shop’s existing business processes. “If you look at the organizations that are outperforming their peers relative to those that are underperforming their peers, the outperformers have a much higher instance of doing business analytics optimization at the enterprise level. It's higher than 8 to 1," said Steve Lavalle, global head of strategy for Big Blue’s BAO arm, in an April interview.

Analytic excellence -- as Lavall himself stressed -- isn’t a turnkey proposition. Thus the exigence for BAO, a composite offering that comprises end-user and analytic back-end software, database and middleware components, and services. It’s in this respect, says Mark Madsen, a principal with BI consultancy Third Nature (and a veteran data warehouse architect), that the SPSS acquisition plays to Big Blue’s strengths.

“Something to think about is how data mining and stats are used and who's doing the work. Normally it's a handful of experts using these tools in the organization. The output is either individual analysis or it's a model which is then built into transactional systems. The latter takes integration and custom work. That high-touch work requires integration software and developers, and IBM offers a lot of both,” says Madsen, who suggests that IBM’s bid should help raise the profile of analytic power users.

“IBM buying SPSS is one more argument that statistical/mathematical knowledge is developing into a key IT skill,” he argues. “I've thought stats and the ability to construct models should be making a comeback in organizations that have been firing these sorts of people for the past decade. Rolling up some numbers in a report is the lowest bar. Deeper insight only comes when you apply more sophisticated techniques.”

There’s a sense in which a brutal economic climate -- or, more to the point, the mistakes, excesses, or transgressions of a financial sector that helped precipitate that climate -- is fueling the demand for cutting-edge analytic technology.

To that end, a host of start-ups are shopping general analytic technology, specific analytic services, industry-specific analytic solutions, or some combination thereof. Consider financial analytic startup Caissa, which develops risk management, quantitative analytic, and portfolio management software. Founded by a passel of Wall Street veterans, it purports to address the very concerns -- modeling, integration, and industry-specific analytic know-how -- Madsen identifies, albeit in a specific vertical niche. Big Blue -- with its vaunted Global Business Services (GBS) arm, its BAO service, its creditable business intelligence and database software, and its best-of-breed data integration middleware -- can trumpet similar expertise across a range of verticals.

The addition of SPSS’ predictive analytic, statistical analysis, and data mining tools amounts to an indisputable -- if at this point not-quite-clearly positioned -- feather in Big Blue’s analytic cap, Madsen indicates.

A Surprisingly Low-Key Approach

IBM itself took a surprisingly low-key approach to the subject of the acquisition.

Analysts who yesterday attended Big Blue’s Smart Analytics conference yesterday in New York said IBM did little to flag what -- if nothing else -- amounts to a significant cash outlay for the company (a purchase that’s comparable to what Business Objects paid for the former Crystal Decisions -- or what Big Blue itself spent to acquire the former Ascential Software).

“They barely talked about SPSS, even though the head honchos [i.e., both Steve Mills and Ambuj Goyal) were there and they had a chance to discuss. At most, they said SPSS was about future opportunities with Smart Analytics,” observes Wayne Eckerson, director of TDWI Research.

Industry veteran Cindi Howson, a principal with BIScorecard.com, says Big Blue’s move builds on a previous (OEM) relationship that it entered into with SPSS nearly two years ago. “IBM Cognos and SPSS had previously signed an OEM agreement but it appeared never to get much traction in terms of deep product integration and appeared to be more marketing related,” she notes.

IBM and other players have been making noise about analytics, Howson says, but driving analytic uptake has proven to be difficult, to say the least, in practice.

“From an IBM perspective, much of the predictive analytics capabilities were largely service engagements. All the vendors have been trying to figure out how to move analytics from a purely ‘backroom task’ to one that is more front office and that makes the results of the models more widely consumable,” she explains, adding that analytic champion SAS has perhaps been most successful in this regard. “SAS is clearly ahead on both points -- the analytics can get pushed to the database for faster execution but surfaced within a Web Report Studio report for wider consumption.”

Elsewhere, she notes, the acquisition could complicate matters for SPSS’ OEM customers -- including IBM rival SAP BusinessObjects. “I will be most interested to see what happens then with the SAP BusinessObjects-SPSS relationship,” she continues. “SAP BusinessObjects launched its Predictive Workbench more than a year ago that allows SPSS Clementine to access universes directly. The whole BI landscape is one of much more co-opetition these days, but I can’t help but recall how the plans to bundle a starter version of BusinessObjects with DB2 Infosphere was dropped after IBM’s acquisition of Cognos.”

Along with SAS, SPSS is one of the biggest names in analytics, statistics, and data mining. It’s considerably smaller than its Cary, N.C.-based competitor, however. Third Nature’s Madsen, for his part, suggests that Big Blue’s stewardship could help significantly broaden SPSS’ market share.

“SPSS had good products and presence. IBM's marketing and budget could take them into SAS territory.”