Information Age: News, analysis & insight for IT & business leaders

"Everyone makes mistakes," says Jack Noonan, chief executive of statistics and predictive analytics software company SPSS. He was referring to his company's mid-March 2004 announcement that it must restate its accounts for the financial years 2001 to 2003 - a move that delayed the company's filing of its 2003 results and precipitated a slew of lawsuits that allege that Noonan and his chief financial officer, Edward Hamburg, made statements that artificially inflated the price of SPSS shares.

Mingling with UK customers at the company's user conference in London last month, Noonan seemed relaxed, as if he had left his troubles in the US. But the restatement is an embarrassing blow for a company that claims its business intelligence tools help customers "draw reliable conclusions about current conditions and future events". Among its key areas of focus: compliance and fraud detection.

For legal reasons, Noonan is unable to say much about the restatements, except that it relates to a change in the way the company is accounting for an acquisition. He is happier talking about a different aspect of SPSS's business: its expansion into analytic applications.

The company has its roots in tools for analysing survey data and its predictive analysis technology enables users to take data describing people's characteristics, attitudes and behaviour and evaluate it using statistical, mathematical and other means. These techniques then generate data models that businesses can use in a number of tasks, including classification, segmentation, forecasting, pattern recognition, sequence and association detection, anomaly identification and profiling. In the past five years, Noonan says, SPSS has concentrated on hiding much of that heavyweight technology behind applications that can be used by a number of different groups with different levels of technical skill, "from the most heavyweight statistician and classic data analysts to marketing and call centre professionals".

In line with that, the company recently launched two key new applications: Predictive Marketing (which helps marketing professionals to determine which customers to send offers to, when to send them and which channels to use); and Predictive CallCenter (an application that integrates with call centres to identify which customers offer opportunities for up-selling, cross-selling and retention deals.)

These new applications are the fruits of SPSS's November 2003 acquisition of Data Distilleries, a small Dutch company, which has brought SPSS a stronger foothold in Europe and an impressive roll call of local customers, including ABN Amro, Fortis Bank and Vodafone. These new products make up only a small, "single-digit" percentage of SPSS's sales, says Noonan, but the company's key focus is on changing that. It can expect stiff competition. Analytic applications are a major focus for business intelligence giants Business Objects and Cognos, to name but two.

The difference between the software offered by these companies and SPSS, says Noonan, is SPSS's focus on prediction. Other analytic applications, he argues, "only look back", since they are engineered to analyse historical data on past transactions. However, most executives will agree that past transactions provide a very good clue to future results. Noonan's focus now must be to make issues surrounding his company's own transactions are a thing of the past.