At Montana State Fund (MSF, Helena), senior executives often spent their valuable time reconciling disparate reports, owing to the limits of the workers' compensation carrier's aging operational database. Today, owing to a multi-year enterprise business intelligence (BI) initiative, the same executives share a "single version of the truth" across about 300 standard reports that bear a data certification "watermark."
Prior to MSF's BI initiative, members of the carrier's executive staff would frequently be confronted with differing numbers to represent some aspect of business performance, according to CIO Al Parisian. "At the end of the discussion MSF president and CEO Laurence A. Hubbard would say, 'Go now and come back with a number you agree on,'" Parisian recalls.
Parisian's boss was frustrated by the diversion of senior executives to duplicative data reconciliation tasks Parisian adds, but was even more concerned about having consistently reliable, precise and accurate information to support vital decision-making.
Parisian's diagnosis of the operational data warehouse's obsolescence dated to shortly after his arrival at MSF in 2005, the year MSF's governance group, the Executive Steering Committee (ESC), was founded. "I could see that our executives were data-centric, but they were working with their own data," he says. "People were could not rely solely on the operational data warehouse and instead relied on unique combinations of manipulated and translated data. Also, data definitions meant different things to different people."
In response to the legacy system's shortcomings, Parisian methodically built a case for the enterprise BI initiative in early 2006 while completing the first major project taken on by the ESC — replacement of the carrier's legacy claims system with Guidewire (San Mateo, Calif.) ClaimCenter.
"I took time to gather many kinds of actions and events caused by the existing operational data warehouse and package them into a case for renovation of our capabilities," Parisian relates. "I approached several executives with examples of decisions that would have been significantly improved by more precise and complete data."
MSF's board of directors approved Parisian's plan, and an RFP was issued in August 2006 for a three-year effort. In March 2007, MSF engaged Millbrook to take advantage of what Parisian characterizes as the consulting firm's insurance data schema expertise.
"Millbrook showed us how to abandon our old business assumptions in favor of a data-centric vision of how we really want to run the business, and what data we need in order to do it," Parisian says. "They essentially posed the question: If you had a blank slate to run a workers' comp insurance company in Montana, what data would you need."
During the first phase of the BI initiative, which ran from June 2007 to June 2008, MSF created and loaded new data schema, including roughly 1600 policy and claims data elements, onto infrastructure based on Oracle (Redwood Shores, Calif.) Business Intelligence Enterprise Edition suite. Following some enhancements in Dec. 2009, MSF went into production with the 1600 data elements in early 2009, according to Parisian. "Three months later, we killed and buried that old data warehouse and weaned the company off it," Parisian adds.
During a second phase of the initiative, which lasted through Jan. 2010, MSF added a further 600 medical data elements, including codes for pharmacy items and medical procedures. The tab for the first phase was about $3 million, and the second phase cost an additional $500,000.
"Using the total of about 2100 elements, we produce about 300 standard reports with about a dozen filters, which are available to every user in the company," says Parisian. "For example, you can pull an agency profitability report at a high level and customize by date range, geography, agency or agent, or by underwriter or class code — all without the intervention of IT."
Users are free to export standard reports into Microsoft Excel for further manipulation, but only the original reports bear a data certification "watermark" that identifies them as authoritative, according to Parisian.
"As long as you stay within that world of 300 or so reports, everything will have a watermark on it," Parisian explains. "If you present a report with that watermark on it, everybody knows that we have scrubbed that data, that it's been audited, it's absolutely up-to-date, all the exclusions are explicit and that report is gospel."
"If users want to manipulate the data further, they can do a mash-up of their own but the moment you download the data, you lose the watermark," Parisian adds. "We're not discouraging that kind of hybrid analysis but we are saying that the easiest path to a compelling argument is to stay within this 'walled garden.'"
As a result of the initiative, executives can access and analyze any of the data elements as they pertain to records going back several years. "They can essentially do extremely enabled business analytics on the fly," Parisian says.
The new BI infrastructure has been very positively received, Parisian says. Not least among its admirers are MSF's president and CEO. "In the year since we've implemented our new BI facility, he hasn't had to send people away to get the right number even once," Parisian notes. "It used to happen perhaps once a month; now it never happens.
Anthony O'Donnell has covered technology in the insurance industry since 2000, when he joined the editorial staff of Insurance & Technology. As an editor and reporter for I&T and the InformationWeek Financial Services of TechWeb he has written on all areas of information ... View Full Bio