From potentially volatile oil terminals in the open ocean to manufacturers in California’s earthquake zone, RLI...
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
insures the otherwise uninsurable. But the insurance firm relies on specialized data analytics software to avoid unnecessary risks.
Based in Peoria, Ill., RLI is a property and casualty insurance firm that focuses on what it calls “underserved” markets. The firm pulls in an average of $600 million in revenue per year, according to Seth Davis, vice president of internal auditing. It is Davis’ job to make sure RLI is collecting every penny it’s due.
Davis runs RLI’s six-person internal audit division. If RLI issues duplicate bills, fails to collect funds from a responsible third party, or if it is the victim of financial fraud by a customer, Davis and his team are tasked with spotting the anomalies. Until 2004, the team manually integrated data from disparate sources and scrubbed the data before moving on to analyzing it.
“That was taking us a couple of weeks out of every month,” Davis said. When the data was finally ready each month, he and his team scoured the data looking for any “money left on the table,” as Davis put it.
The problem was that RLI had no way to identify which clients posed the most risk for fraud or billing mistakes. That meant Davis and his team had to either analyze all of RLI’s customer data -- a time-consuming endeavor -- or focus on just a sampling of the data and risk missing something.
Popular with insurers and accountants, CAAT software lets auditors create custom scripts to integrate data from multiple sources, then applies specialized filters to the data to identify potential fraud or fund leakage.
CAAT software is also used by government tax agencies, including the Internal Revenue Service and the Massachusetts Department of Revenue, to review tax rolls and catch delinquent tax payers.
At RLI, now rather than manually reviewing all of its customer and financial data, CAAT software applies exception-based modeling techniques to flag just the most likely candidates for fraud or other billing errors and weighs the results. That way Davis and his team can focus on the highest-risk cases.
Thanks to CAAT, RLI’s audit team has cut the amount of time it spends reviewing potential fraud and fund leakage cases by 25% even as it has increased the amount of fraud and fund leakage it detects, according to Davis.
On average, RLI now catches $100,000 in fund leakage on an ongoing basis. In one case, the company discovered an error in a query that had resulted in $4 million in missed billings to multiple clients, Davis said.*
But Davis gets as much pleasure finding money left on the table with data analytics as he does not finding it. A good audit, he said, doesn’t just identify fraud or missed billing opportunities but provides assurance that the company is operating efficiently when it doesn’t find any.
As for the software itself, Davis said it has proved fairly easy to use, requiring only a few days training for new audit employees. Davis himself has no IT background but is comfortable using the ACL suite.
“We’re not IT people,” Davis said of his team. “This was all done by a couple of guys that didn’t know how to spell IT,” he said, adding, “You don’t have to have that technology background to use a tool like this, to be effective.”
What proved more challenging, he said, was developing an understanding of how all the different types of data involved in an audit relate to one another, a job that technology alone can’t help with.
Davis also urged companies using CAAT to regularly look for ways to update the data analytics models to stay up to speed with changing market conditions.
“You can’t just create these scripts … and never enhance them,” Davis said. RLI holds quarterly meetings to brainstorm enhancements to its data models, discuss how to take advantage of new functionality in the software, and generally just “become more efficient in the analysis.”
* Due to a reporting error, the original version of this article misstated a technical term related to RLI's business. It has been updated to reflect the correct industry terminology.