While analyzing customer relationships is key to consumer-facing industries such as retail, businesses that sell to other businesses are increasingly using sophisticated analytics as well.
Whether its reining in costs to improve efficiency or finding valuable information in data, business-to-business companies are turning to advanced data and text mining, as well as analytical CRM (customer relationship management) and business metrics.
Niis/Apex Group Holdings, which provides risk management and reinsurance services to health insurance and self-insured companies, is using analytics to help those businesses better manage spiraling health care costs and promote wellness among patients.
Niis/Apex uses data mining technology from SAS Institute Inc. to accomplish this. Its in the second stage of a three-stage process, according to Jody Porrazzo, director of econometric risk strategy at Apex. “Our goal is to get our actuaries closer to the data,” said Porrazzo, adding that by using SAS Enterprise Guide application, Niis/Apex actuaries can easily and quickly access data without IT assistance, a process that might have taken days before.
Having a solution in place that gives everyone access to data was the first phase of Porrazzos project. The current phase includes adding what Porrazzo calls heavy-duty predictive analytics, trending and time-series analysis.
In Phase 3, Apex plans to adopt SAS Enterprise Miner and Text Miner. Enterprise Miner adds a graphical interface to the data mining process, including if … then scenarios, while Text Miner adds analysis of textual data such as symptom reports and nurses notes. “Phase 3—thats Star Wars, thats The Matrix,” said Porrazzo in Princeton, N.J.
Both rising health care costs and patients lives are at issue, according to Porrazzo. Through SAS analytics, actuaries can track correlations between the use of fertility drugs and the incidence of babies born with low birth weight or compare the cost of using experimental drugs to reduce brain swelling caused by head trauma with the cost of continuing care for brain-injured patients.
The applications also look for indications that patients could be at risk for requiring expensive medical treatment and look to prevent traumatic injuries, diabetic shock and chronic heart disease, long before such treatments are required.
“The most important thing is disease management,” said Porrazzo. “We need to help customers plan and budget for future costs and prevent something bad happening to patients whether its three years, 25 years or 50 years down the road. Most importantly, its about keeping people well.”
The system can analyze patient medical records and spot risk factors for chronic heart failure, such as smoking, obesity or high blood pressure, said Porrazzo. Ultimately, she said, it will take text mining to make the analysis complete.
“We have access to structured data now; we can put it into a table. But we need to be able to look at unstructured data,” said Porrazzo. “Theres some very important information in that unstructured data. Thats the future of where my company needs to be.”
According to Porrazzo, nothing can match the high-powered analytics that SAS provides. “Most predictive models out on the market use canned regressions, t-tests and correlations. Those arent very robust models,” she said.
Next Page: Fuzzy Clusters and Neural Networks
Fuzzy Clusters and Neural
Networks”>
SAS, on the other hand, supports more advanced data mining, including fuzzy cluster data models and neural networks, that help Apex meet its business goals, according to Porrazzo.
“We need to lower costs [and] have positive outcomes. We need to manage high-cost claims by identifying those patients that are potential high-cost claims,” Porrazzo said.
B2B analytics was a key part of the recent outsourcing deal IBM announced with Dun & Bradstreet Corp. last month. D&B, of Short Hills, N.J., is outsourcing its customer service, data acquisition, and delivery and financial processes to IBM, of Armonk, N.Y. The deal calls for IBM to apply real-time analytics to the more than 100,000 D&B client relationships in a bid to improve client service and reduce costs.
And The CEI Group Inc. uses business analytics for applying risk management to the vehicle and driver management services it provides to corporate clients, including processing accident claims and managing vehicle repairs.
CEI uses Business Objects S.A.s business intelligence platform technology to track data generated by CEIs custom workflow application known as IntelliClaim. By tracking more than 100 metrics daily, CEI can measure timing, cost and control for nearly every component of its vehicle repair and safety programs.
With Business Objects-based technology, CEI can determine what drivers are at high risk for accidents and what vehicles are responsible for the most claims. The system can also determine accident rates and repair costs for vehicles, ultimately affecting CEIs future fleet purchasing decisions, said company officials in Trevose, Pa.