The application of predictive analytics to underwriting often has been cast as a question of art versus science. Traditionally, underwriters have shown some resistance to the penetration of technology into their territory on the grounds that, when all is said and done, underwriting is an art. But more recently, they seem to have acknowledged that rather than being adversaries, science and art are partners in a common enterprise.
Underwriters have come to appreciate technology much the way agents have: They have discovered that it makes their lives easier, freeing them of administrative drudgery and conserving their time for tasks that genuinely require their intelligence and expertise.
A further driver of the penetration of analytics into underwriting and pricing is the inexorable demographic fact that underwriters are becoming scarce. With the number of underwriters diminishing, technology simply is necessary to pick up the slack. That has been the case in the increasing development of underwriting support systems that use rules engines and analytics to "jet issue" or otherwise simplify the underwriting process by automating those aspects of it that -- with the aid of technology -- can literally be called "no-brainers."
The utility of analytic technologies for underwriting and pricing, however, goes well beyond the efficiency gains. More and more, analytics are being applied to create a competitive edge in an increasingly commoditized business, both at the micro level of pricing and at the macro level of product-portfolio and line-of-business performance.
"We're taking this to the next level of sophistication in terms of identifying the analytics that will generate truly unique insights for building more competitive products because that's really where the game in insurance will be won," opines Gail McGiffen, a partner with Accenture (Chicago). Unfortunately, carriers anxious to improve their competitive position all too often have embarked on data capture efforts that have yielded more quantity than quality, she suggests.
As impressive an accomplishment as capturing more data more efficiently and organizing it neatly into data warehouses may be, insurers need to focus on precisely what they are going to use that data for, McGiffen says. "It isn't enough to just dump data into management information systems and reports," she insists. "It is extremely important to focus on the KPIs [key performance indicators] that drive business outcomes."
Other insurers would do well, McGiffen advises, to heed the concerns expressed by an Accenture client: "The last thing in the world I want is to have my very busy underwriters in the field getting more reports with all this data loaded in their in-boxes," the client said. "Where I really want to go is faster and, yes, more reports, but greater precision in the types of analytics and therefore the data to support it."
When carriers do approach predictive analytics build-out with the proper focus on quality, the implementation process itself is likely to be illuminating, according to Mark Gorman, the Minneapolis-based principal of consulting firm Mark B. Gorman & Associates. "Typically the communication they share and the insight they gain have already improved the business even before the implementation," he comments. "This is all about turning data into information, information into knowledge and knowledge into behavior -- whether that behavior be human or automated."
In the P&C industry, some of the greatest analytics-related focus currently is on a combination of human and automated behavior. As carriers have mastered the use of rules and analytics for underwriting simpler lines of business, such as personal auto, P&C insurers have begun to focus more on applying predictive analytics to automate some commercial lines, such as commercial auto and business owners' policy (BOP), according to Gorman.
Complex commercial risk -- along with non-P&C risks analogous in complexity -- is not amenable to the "jet-issue" treatment, Gorman adds. But it is very much amenable to the application of analytics in the form of automated decision-support rather than decision making. Essentially, Gorman says, "The system can make a recommendation that the underwriter can override, ignore or make an exception to."
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