The Historic ChallengeInsurers have historically invested significant time and money on projects designed to improve data management and quality. Create a single source of the truth; have the right data, in the right place at the right time; provide data quality improvements; and, establish stronger analytics were among the rallying cries. While these objectives were frequently associated with data warehousing initiatives, major data initiatives equally applied to the implementation of policy processing, claims, billing systems, and portal solutions. The vision of the future state was easy to grasp, but a key measure of success involved addressing the underlying data requirements. Months were spent creating data dictionaries and data models that reflected the needs of the organization.
Part of the challenge was developing an industry-specific model that addressed required data elements and their relationships, supporting both business function and product consideration. This was no simple feat. Endless meetings defining, debating, and refining the models would occur. Time would pass, and the vision would fade. The cost would greatly exceed expectations.
Some large carriers would engage IBM for access to their Insurance Application Architecture (IAA) and, ultimately, their reference models. Others would look to software companies or consulting firms for support. For example, many consultancies have focused on developing and refining industry Business Process and Data Reference Models in the verticals they serve. These models would function as a communication tool during their engagements. In general, these models were useful for the organizations that had them, but they were neither comprehensive (with the exception of IAA), nor were they supported by other organizations like software providers, which would bring a broader benefit.
Why no standard model? A key reason was that having a well-developed model was often considered a competitive advantage for a carrier or a marketing tool for a services company. This belief reduced the desire of those organizations to support the development of an industry-wide model. For many years, companies had little choice but to attempt to develop their own from scratch. Today, in an era of open source computing and widespread collaboration, views have shifted, and it is recognized by many that having a common model is ultimately beneficial. It is understood that the model has value, but it is not a competitive advantage in itself.
Insurance Data StandardsThere has been a long history of establishing standards in the insurance industry. Common forms, statutory reporting standards, EDI standards, etc. have long been supported by organizations such as ACORD. Historically though, the focus has been on the exchange of content between organizations. These standards became part of the business requirements that IT needed to support, but there was little internal value within the enterprise.
Since the middle of the last decade, there has been a focus in the development of industry standard process and data models that support the internal needs of insurance carriers. The two leading initiatives underway are being supported by ACORD and the Object Management Group (OMG). Both of these initiatives owe a debt of gratitude to IBM, which donated components from their IAA model to the cause.
OMG has a long history of providing support, management, and distribution of industry-wide standards in Data Modeling, Business Process Modeling, and Service Oriented Architecture. Advancement of Unified Modeling Language (UML), Business Process Modeling Notation (BPMN), Model Driven Architecture (MDA), Common Object Request Broker Architecture (CORBA), as well as thought leadership in Enterprise Architecture and Service Oriented Architecture are some of the key contributions OMG has made to the technology world. The Insurance Work Group falls under their well-established Finance Domain Task Force. The Insurance work group represents a wide range of carriers, product companies, and service organizations.
The OMG Insurance Working GroupThe stated goal of the Insurance Working Group is to develop a Standard Reference Model for the P&C; insurance industry. Ultimately, these models and their supporting artifacts can be reused and/or extended to other related domains like life and reinsurance.
The deliverables for the work group include:
- Glossary of P&C; Business Terms and accompanying metadata (source, version, format as well as party, party role, address for example).
- Conceptual Data Model representing P&C; business concepts.
- Attributed Logical (ER) Data Model with appropriate identifiers, logical data types and relationships as appropriate.
- Traceability of P&C; Business Terms to models and model elements listed above.
The model produced by the OMG initiative will be widely published and can be levered by enterprise and solution architects, data architects, data modelers, and business analysts during their projects. Several member organizations are using the current draft of the model on their internal projects and providing feedback to the group, allowing the models to be updated based on carrier experience.
Subscribing to an industry standard data model like the one being promoted by OMG can provide significant value during a wide variety of warehousing, business intelligence, and data migration projects. However, a bigger opportunity exists in how these models can be leveraged to improve the use of data and in support of data governance within the enterprise. This is a topic to be explored in future publications.
Additional information is available on the OMG Insurance Working Group Wiki.