An information model that precisely defines insurance-industryconcepts and the relationships among them—a key element in datastandardization for insurers—is at hand.

Beta software versions of the model have been made available inrecent months by ACORD, which is expecting to release a finalversion of the model soon—perhaps before the ACORD LOMA InsuranceSystems Forum concludes.

In addition to collaborating with ACORD on finalizing the modelover the past year, IBM has been demonstrating to insurers thebusiness solutions the company has developed based on its own work,beyond its partnership with ACORD. The IBM Insurance IndustryFramework addresses core insurance, distribution, risk management,and customer care and insight.

Two IBM experts in the field are scheduled to update conferenceattendees on the information model, IBM’s solutions and thepotential benefits of it all during an opening-day sessiontoday.

The scheduled presenters are Edith Pula, an industry-modelssales executive for North America in the Information Managementsection of IBM’s Software Group; and Oliver Koernig, aninsurance-solution architect in the Software Group’s IndustrySolutions section.

Their session, “IBM’s Insurance Industry Model And ACORD: AnUpdate On Progress,” is scheduled for today at 12 noon.

The information model is key to the ACORD Framework, which alsoincludes four other interrelated models. All are designed for usein developing data standards that are not dependent on insurers’products, lines of business or geography.

The four interrelated models are:

• A business dictionary, whichprovides standardized definitions of insurance concepts—such aspolicy, product, party and claims—as well as synonyms,business-line specific usage and references.

• A capability model withprocess maps. This model identifies the baseline activities thatinsurers need to perform and includes a list of process maps forsome of those capabilities. This model can be used in processengineering, evaluating mergers and divestitures, and analyzingbusiness operations.

• A component model, whichconsists of a set of reusable components for various data services.The model, organized around the concepts that other modelsdescribe, can be used in rationalizing and designing aninformation-technology portfolio and as the basis in building orbuying software.

• A data model, which includescreating data warehouses and validating a company’s own datamodels.

The information model, which has been developed with the help ofseveral pieces of intellectual property that IBM has donated to theproject, defines the relationships among all of the variousinsurance concepts.

The information model is critical in establishing clear lines ofcommunication between insurance-company executives andtechnologists, Pula and Koernig explain. Executives want data toolsthat help them analyze how operations are performing and theinterdependencies among operations. The technologists areresponsible for providing those tools.

But understanding which information is necessary to build thosetools “is where the transaction can break down” without the aid ofthe information model, Pula says.

A real-life example illustrates the issue, Pula says: Executivesfor a long-term-care insurer wanted their technologist to build animproved claims-processing tool. At the onset of the project, twoclaims processors described their duties for the technologist. Oneprocessor was far too detailed, and the other too vague, so neitherwas helpful. But the IBM models provided the technologist the rightamount of detail to begin building the tool, Pula says.

Besides continuing to work with ACORD on the information model,IBM in the past year has been demonstrating to insurers how thecompany’s Industry Models for Insurance could help carriers’improve operations and develop products. The models define nearly300 assemblies of data in describing the information that is neededto analyze, among other areas, claims, new business, financialreporting, risk management, underwriting, pricing andoperations.

For example, while many tools and applications are available toassist insurers in predicting claims and identifying fraud, theyare ineffective without the appropriate data, Pula notes. But datarepositories based on the IBM Industry Models for Insurance helpdefine what information insurers need to use the advancedanalytics.

Modeling also would allow insurers to improve their holisticunderstanding of policyholders, according to Pula and Koernig.

“Understanding a customer holistically has high value to aninsurer,” Pula says.

For example, many insurers do not recognize when a policyholderhas multiple coverages and what opportunities or risks thatcreates.

“If I'm creating a marketing campaign and I know that a ‘mere’renter’s policyholder also owns two other homes, has children andowns two vehicles, I can offer that seemingly low-value customer agreat cross-sell deal,” she says. “If I know that they arealready a customer in my other lines, I can pass a valuable lead tothe agent or broker to offer a full financial assessment to fillany insurance gaps.”

This base of knowledge also can work “conversely,” Pula says. “Icould find out that I have too much exposure with a particularcustomer and take steps to mitigate it.”

It also could have claims-handling implications. “I have abetter understanding of my entire relationship with thatindividual. I can identify conflicts of interest. I can service aclaim with more comprehension of the claimant's total involvementwith me and value to me. An insured who is also among my top-10producing commercial general agents may warrant a different levelof customer service,” Pula says.

For example, Pula recounts one insurer’s experience incancelling the auto policy of the CEO for a major corporate clientdue to excessive claims. “There was no way that this individual'spersonal and commercial relationships were understood together,”she says.

“When the auto cancellation came, the CEO was so upset that hecancelled the commercial insurance. Understanding this could haveprevented a rote transaction from becoming a major commercial lossor at least had given the carrier the chance to assess thepotential damage beforehand and decide what action to take or howto mitigate the likely response.”