If SAS Institute CEO Jim Goodnight is feeling pressured by IBM's recent purchase of rival predictive analytics vendor SPSS, he's not showing it.
"I haven't noticed much difference, really, since IBM bought them," he said in an interview Tuesday at SAS' campus. "We have competed with SPSS for 35 years. We've competed with IBM that same amount of time. Nothing really new competitively has taken place as far as we're concerned."
SAS bills itself as the world's largest privately held software company, reporting $2.26 billion (£1.38bn) in sales during 2008. It has seen heightened competition of late, as vendors such as IBM, Oracle and SAP aggressively flesh out their BI (business intelligence) strategies.
IBM's SPSS purchase particularly seemed aimed at SAS, given the latter's long track record in predictive analytics, which centre on modelling future outcomes and conducting "what if" scenarios, rather than generating reports from historical data stores. Forrester Research has predicted further consolidation in the predictive analytics space, which includes a range of smaller vendors as well as open source projects, in the wake of IBM's move.
IBM is hoping to boost predictive analytics projects with a new 4,000-strong services arm it formed earlier this year.
"I hope they can find work for all of them," Goodnight said. "Analytics is big, it's certainly the next wave, but I'm not sure it can grow as rapidly as IBM thinks it can."
One SAS package is supposed to help debt collection operations. It can be used to develop customer models showing which means of contact is best for various types of debtors. For example, companies could use it to make sure live collections agents focus on people who are the most likely to actually pay back some money, given the expense of running a call centre.
The National Hockey League's Carolina Hurricanes are using another "optimisation" software application from SAS to determine how to price game tickets, said Bill Nowicki, director of ticket operations, during a panel discussion at SAS headquarters.
"On an annual basis, our executive team gets together and tries to determine, based on the previous sales cycle, what the optimal base price will be," he said. "[We've looked] at our promotions we've run previously, and seen how well they fared." But the team lacked a "scientific model" that could look at those past sales, analyse them and come up with a price that would boost sales but also keep the team competitive with other entertainment options, he said.
"We had to figure out a better way to do that than gut instinct," he added.
SAS is also planning to discuss a new application, SAS Enterprise Case Management, which helps fraud investigators document ongoing cases and assists with making required regulatory filings.
Companies should use technology to systematically detect fraud because "fraudsters" behave like a network, said Chris Swecker, a former assistant director of the Federal Bureau of Investigation's criminal investigative division and corporate security director for Bank of America, who now works as an independent consultant. He also spoke during a panel discussion.
By addressing fraud in a "one off" fashion, companies end up playing a frustrating game of "whack a mole," he said, referring to the carnival game where players try to smack a plastic mole with a mallet as it rapidly and randomly pops out of holes.
Another persistent problem is that few fraudsters ever get prosecuted, he said. The FBI transferred 2,500 agents to counterterrorism efforts some years back, he said. "That's a lot of agents when you consider there's only 11,000. That meant we had to set higher thresholds. We would not investigate a [fraud] case under $500,000. So, most of these singular fraud cases will never get prosecuted. ... The universe of fraudsters out there is not diminishing. It's a very low risk environment for them to operate."
Find your next job with computerworld UK jobs