SAS Institute on Thursday announced a new toolset aimed at giving business users the ability to work with predictive analytics software, which has historically been the province of specialised statisticians.
Predictive analytics refers to the practice of crunching existing data sets in an effort to determine patterns and predict future outcomes.
SAS' Rapid Predictive Modeler tool will enable business users to analyze scenarios like customer segmentation and customer "churn," or turnover, through a set of prebuilt models, SAS said. Results come out in easily understandable reports and charts.
Models generated with the tool can be further tweaked with SAS' high-end Enterprise Miner platform, and also can be run on Teradata, Netezza and IBM DB/2 databases.
The tool is now generally available worldwide, but pricing information could not immediately be obtained Thursday. Current Enterprise Miner users will get the new tool as part of version 6.2, SAS said.
SAS' announcement reflects a desire by customers to get more people using predictive analytics tools, said IDC analyst Dan Vesset. But "the key message is who the expanded user base will be," he said. "You still need to be a business analyst type, not a basic marketing person."
In turn, the tool presents a trade-off between ease of use and the type of flexibility statisticians may desire, he said. "This on the one hand allows an end-user to work with a defined data set they understand. But if you want to really experiment, test and build your own models, then you'd go back to Enterprise Miner."
The market for predictive analytics is "pretty healthy," albeit dominated by SAS and IBM's SPSS division, but products like Rapid Predictive Modeler have the potential to increase its growth rate, Vesset said.
There is also growing interest in the open-source R language for predictive modeling, which has its roots in academia. Commercial companies such as Revolution Analytics have been popping up, offering software and services for working with R.