Optimizing data and cash are significant mandates for tier one institutions, as banks weather the remaining recession and position for recovery. IBM's offer to buy Chicago-based predictive analytics giant SPSS for $1.2 billion reveals Big Blue's efforts to gain share by helping banks save money while making the most of the endless data pouring into banks' marketing, risk and fraud operations.
"At the highest level what we know is in all industries the amount of information is growing, and continues to grow at an exponential rate," says Eric Yau, vp of business intelligence and performance management at IBM. "And of course we know data does not equal useful information."
IBM's investment in predictive analytics is timely, with business intelligence, data and analytics, "probably the top three areas of focus in the new competitive world," says Omer Sohail, head of Accenture's information management services for the financial industry.
Many banks are re-examining how they aggregate data and what kinds of BI tools they use - but in this economic environment, it's primarily from a cost perspective. "There is so much focus on cost reduction, the idea is to do a BI rationalization of your tools," Sohail says.
IBM's plan to buy SPSS is the latest deal in a business analytics and intelligence market that's been hot for several years, marked recently by EMC's bidding war with NetApp for DataDomain. Other buyers in the market over the past two years include Oracle and SAP, which have all increased their focus on business intelligence.
Of all the players, IBM has opened its wallet widest. The company offered $50 a share for SPSS, a 42 percent premium on the stock's closing price the day the deal was announced, on top of $10 billion other analytics buys since 2006. With the deal, IBM will form a new, 4,000-employee strong service line, Business Analytics Optimization, the first new line launched by IBM Global Business Services since the division was created in 2002. If successfully executed, the deal would give IBM heft in the combined space of business intelligence and predictive analytics, perhaps enough to counter the dominance of SAS, which owns more than 30 percent of the market.
IBM needs SPSS to pull off this competitive move, because though IBM's got a good play in business intelligence, its predictive capabilities have been weak. This is where SAS differentiates itself, noting the predictive nature of what it dubs "business analytics" allows companies to focus on defining causality, forecasting and optimization.
"I give a lot of credit to Cognos, Business Objects and MicroStrategy for creating a lot of noise around business intelligence as something that adds value in itself," says Russ Cobb, vp, alliances and product marketing at SAS. "But doing queries to find out what happened in the past doesn't help a whole lot in decision making."
Forrester Research analyst Boris Evelson agrees, but says IBM's marriage with SPSS is a step toward closing that gap. With an IBM/SPSS combination, the "BI world moved a step closer to addressing one of the gaps in traditional BI product lines - integrated advanced analytics," Evelson writes, adding "IBM can compete more effectively with SAS, TIBCO Spotfire, MicroStrategy, Microsoft and Information Builders in deals that require both traditional and advanced analytics."
Banks adopting predictive analytics and optimization find the technology enables a paradigm shift in customer engagement. At First Tennessee Bank, SPSS's PASW modeler has been put to work helping the bank transform its customer marketing efforts from a "bucket strategy" into a customer behavior-driven segmentation model, says Tanner Mueller, database marketing team manager at First Tennessee.
Dan Marks, the bank's CMO, says the institution has used predictive analytics in its direct marketing efforts for nearly five years, but of late has put it to work institution-wide. By applying predictive analytics to the company's enterprise data warehouse, First Tennessee is able to optimize its marketing efforts, determining which channel is the most cost-effective way to reach out to each customer in its cross sell efforts - whether it be direct mail, email, phone or even the ATM. A cross sell effort to existing checking account customers has resulted in a seven-to-one revenue-to-cost ratio in checking cross sales.
The use of predictive analytics allows First Tennessee to combine propensity models and response models, but also to close the loop by offering timely analytics on the success of campaigns. "But the next frontier is really using this tool set to model and optimize the entire marketing mix," Marks says.
Analysts predict future offerings will include memory analytics, text analytics, process analytics and more - with benefits for marketing, fraud, credit and operational risk. "The goal of analytics is to bring together all of the data and help turn it into more actionable information," Yau says.