[Skip Global Navigation]

 

SPSS tea leaves read the future

Welcome to SPSS South Africa Home

CEO Noonan says company analytics takes a lot of the guesswork out of predicting

October 2, 2007

Tomorrow is always a day away. But today more than ever, business is straining its eyesight to see how people might react to future moves.

Is there such a thing as a workable crystal ball?

Yes, in fact there is. It's the young science of predictive analytic software, and its pioneer developer is right here in Chicago. Originating in 1968 and refining its complex set of algorithms over 40 years, SPSS Inc. (Nasdaq: SPSS) has emerged from under the radar as the only pure play in this billion-dollar market.

In the process, it has become the largest surviving publicly held software company in Chicago.

Predictive Analytics, SPSS CEO Jack Noonan explains, is the mining of data about people -- based on empirical evidence -- and forecasting how they'll act tomorrow.

"We focus on how people think, who they are, their attitudes and what they're likely to do next," Noonan, 60, says.

A veteran systems engineer at IBM and Amdahl before taking over SPSS in 1992, he nurtured its specialty into a full-blown industry.

"Whether sifting through an organization's own data from experience with customers, employees, students, policyholders or other constituencies, PA can accurately predict future behavior," he says. "This has found growing application in cross-sale marketing, fraud detection, recruiting and the mushrooming market for CRM, customer relations management."

SPSS today serves a record 250,000 clients, 60 percent of whom are in 61 other nations, covered by 185 specialists in 12 offices worldwide. Some of them apply the software to identify which customers are the most susceptible to a product upgrade or supplemental purchase and on which customers sales personnel shouldn't waste their time.

Other clients include insurance companies using experiential data to predict which damage claims might have the greatest propensity to commit fraud. Others are universities seeking to predict which students have the highest likelihood of dropping out, so that preventive attention can be given.

You might call it all benevolent profiling.

"All this differs greatly from the traditional science of business intelligence," Noonan says, "which customarily sifts data to report what happened yesterday. We want to understand tomorrow."

SPSS, which is short for Statistic Packages for the Social Studies, warms up its crystal ball for clients such as Cablecom GmbH, Switzerland's largest cable network. Like other broadcast systems it is constantly fighting churn.

"Seeing the world through the customers' eyes is the heart of my business," Cablecom marketing officer Federico Casconi says. SPSS technology crunches the data of his customer surveys plus other factors, and is able to predict with 78 percent accuracy which customers will bolt, enabling the network to act quickly and proactively.

Thus far, he reports, churn rates have been reduced to 2 percent from 19.

Noonan sees his predictive analytics market growing at an 11 percent annual clip.

"Our challenge is to go far deeper within each client organization," he says. "While some deploy our software for all their possible needs, many use it for less than 10 percent of its potential uses."

To aid the penetration and be more useful, SPSS this summer unveiled new versions of its data mining, text mining and survey analytic products, providing a next-generation series with improved integration, automation and information management. It's partly the result of the more than $110 million Noonan has invested in 2005 and 2006 at his four R&D labs in Chicago, Rochester, China and the U.K. He's spent another $32 million in just the past six months.

Following a record second quarter, Noonan says, revenues and operating income continue strong, and he expects full-year revenues to reach $285 million to $295 million. He's increasing his earlier expectations for per-share earnings to the $1.53-$1.63 range before a $3 million restructuring charge, which he hopes to fully recover in 2008.

If Jack Noonan's forecasting prowess matches that of his wares, the company will rack up another all-time high.

Ted Pincus is an adjunct professor at DePaul and managing partner of Stevens Gould Pincus, merger and management consultants.