SPSS Predictive Web Analytics

SPSS is marrying its Web-analytics and predictive-modeling technologies with the launch of Predictive Web Analytics. Designed to vest organizations with the power to convert their Web data into deep customer intelligence, the new application promises to provide historical and predictive insights to help increase value for customers and raise their returns on investment.

“By embedding its Clementine predictive workbench into its NetGenesis Web analytics software, SPSS has taken what we believe is the latest step in the evolution of Web analytics — creating customer awareness and intimacy,” Robert Moran, Aberdeen Group vice president of enterprise business applications and analytics, told CRM Buyer.

Whereas most businesses are using Web analytics largely for low-level functions, such as tallying the number of users, visits or page impressions at a site, or for business awareness to track visitor behavior, SPSS’ new offering enables companies with a Web presence to develop business processes around “how customers respond as individuals rather than just page clicks,” Moran said. “Nobody else does this to this degree.”

Taking It Personally

“Only when a company realizes that the Web site is an electronic persona that reflects its entire being — e.g., marketing, sales — does it start to profile users and content as a way to understand their complex interactions,” according to a recent Aberdeen report.

The “analytic trick” that SPSS has accomplished is to discover that different communities usually visit various parts of the same site from different perspectives, different entry points and at different times, the report said. “Online visitors do not click in unison,” so a more personal approach is called for, the report stated.

SPSS’ Predictive Web Analytics focuses on segmenting which groups of visitors are coming to a site, detecting their product or content preferences, automatically identifying the most significant paths taken through the site and then drawing predictions based on all of this data, Jay Henderson, SPSS senior product manager for Web analytics, told CRM Buyer.

Making It Official

SPSS’ customers have been combining the company’s Web analytics and predictive modeling capabilities for some time, Henderson said. The new offering simply prepackages those capabilities with the aim of “automating the discovery of the significance” of various data, Henderson said. “Lots of organizations face the challenge of information overload and the challenge of user identification.”

The Financial Times’ FT.com is one site that recently began using SPSS Predictive Web Analytics to develop targeted marketing messages to increase site use, retain customers and help anticipate customer behavior regarding subscription renewal. For example, the Financial Times will examine how subscribers have been using the site in the last three months, identifying various patterns of use and who is likely or unlikely to renew their subscription, according to Chris Catchpole, head of database services at the company. One goal is to give users guidance on some site functions they may not be using so that they can “get more value from their subscriptions,” he said.

In the future, the Financial Times plans to look more closely at users’ propensity to make purchases online so that marketing efforts can be targeted based on how individuals or clusters of individuals use the site, Catchpole added.

Pricing for SPSS Predictive Web Analytics starts at US$135,000 plus services. Implementation services are charged on a time-and-materials basis.