Cognos/SPSS Coupling Aims to Fast-Track Predictive Analytics
Partners say new integration between Cognos 8 and Clementine can help accelerate predictive analytic deployments
- By Stephen Swoyer
- March 12, 2008
Predictive analytic technology has been the Next Big Thing for so long that it's in danger of being eclipsed. What's needed, proponents say, is a slam dunk value proposition: streamlined implementation, for starters, coupled with user-friendliness or even user self-serviceability.
That's just what Cognos -- an IBM company (and a segment within Big Blue's overarching Information Management division) -- and SPSS Inc. claim they're offering: a way for organizations to fast-track predictive analytic implementations, in this case, by coupling SPSS' Clementine predictive analytic technology with the Cognos 8 business intelligence (BI) and performance management (PM) stack.
Cognos and SPSS announced the new partnership -- which both vendors bill as a strategic accord -- on Monday, citing the immediate availability of at least one key integration point, along with the expected availability later this year of industry-specific joint Cognos and SPSS solutions.
Right now, users can invoke Clementine from within the Cognos 8 environment, the partners say. Over the next year, the companies plan to deliver targeted predictive-analytic solutions for several verticals, including financial services and insurance (risk management), retail and manufacturing (campaign effectiveness and product placement), and life sciences (sales effectiveness).
"What we're announcing … is a deeper technology alliance with SPSS, where we're going to work together to create leading performance management solutions between Cognos and SPSS," said Jennifer Francis, vice-president of market development with Cognos. "It's not an embedding relationship. This is a go-to-market relationship creating joint solutions with SPSS."
Embedding relationship or not, Cognos and SPSS are delivering canned integration, in the form of a "Score" button, between Cognos 8 and Clementine.
"There's a couple of ways to be able to integrate Cognos 8 and Clementine. If I'm in Cognos 8, I'm a Cognos 8 user, I might want to press the 'Score' button and get back scores for the group of customers that I'm analyzing. [In this case,] Cognos 8 can call back in the background [to Clementine] and get the results," Francis explained. "From an end-user perspective, this is entirely seamless."
That's available today. Over the next year, she continues, Cognos and SPSS plan to execute on a more ambitious go-to market strategy. "That capability is there [today], and what we're going to be building with SPSS are the vertical solutions, the industry models around insurance. [For example,] if I want to score how likely you are to be an insurance risk, there's going to be a Clementine model around that, so we're going to prepackage these solutions for key industries and key solution areas. The idea is to make it faster and easier for our customers to get to value [with predictive analytics]."
It's a compelling proposition, but does it address the most salient customer pain points with respect to predictive analytic implementations? In a recent study from TDWI Research, for example, customers cited a handful of daunting challenges associated with deployments of predictive analytic technologies, such as developing (and sticking to) a multi-stage process (as well as the more immediate problem of codifying a clear project definition), not to mention the half imaginative-artistic, half logical-scientific task of actually building predictive models, for which the hiring of business-savvy analysts is seen as a must.
"[M]ost experts agree that predictive analytics requires great skill -- and some go so far as to suggest that there is an artistic and highly creative side to creating models -- most would never venture forth without a clear methodology to guide their work," writes TDWI director of research Wayne Eckerson.
In the case of their vertical-specific solutions, Francis claimed, Cognos and SPSS propose to do a lot of the pre-implementation heavy-lifting for prospective customers. When they become available, she said, the combined Cognos/SPSS solutions for financial services, retail, and life sciences will include pre-built scoring features that are specific to each industry. In their joint marketing efforts, Cognos and SPSS also plan to offer implementation expertise.
The offering could be a slam dunk for joint Cognos or SPSS customers, Francis argued. IBM itself has a pre-existing relationship with SPSS -- it signed a technology partnership with SPSS late last year -- and customers who use any of Big Blue's Information Management offerings (including the ubiquitous DB2 database) could also benefit from the accord.
"Overall, the Clementine product enables [SPSS'] customers and our joint customers to easily access, prepare, and integrate structured data and survey data into predictive analytic modeling, and it's also well-integrated with IBM's Intelligent Miner and DB2. When you add the [Cognos 8] integration, it's just a very good overall solution for IBM customers and [Cognos/SPSS] joint customers," she concluded.