|
By Peter A. Marotta, AIDM, FIDM, Enterprise Data Administrator, ISO
The age of analytics has shed new light on data as an asset and on data management as the discipline for managing that asset. To manage information resources, data managers have developed a number of best practices over the years. They range from high-level organizational practices, such as data stewardship, to operational processes, such as data standards and data-management tools.
However, for best practices to remain superlative, they need to continually evolve or risk devolving into just another set of common activities without measureable benefits. This is especially true if the related subject matter, in this case data and data usage in the insurance industry, is evolving at an equal, if not greater, pace.
A major countrywide automobile insurance company recently announced it has created an iPhone application to assist its policyholders with the reporting and documenting of automobile claims. This “app” can record via GPS the exact location of the accident, store accident photos and find towing services.
The Mississippi Department of Insurance will be using Twitter to notify residents of impending storms and evacuation information. Insurers across the globe are looking at telematics — the synthesis of GPS, communications technology and automotive navigation systems — for underwriting and pricing.
The breadth and depth of data is rapidly increasing. The number of external data sources — both public and private — is growing, as is the granularity of data. New technologies allow for improved access to information and expanded capabilities of analytical tools. Organizations are making increased use of data, technology and analytics to gain competitive advantages. These advantages include risk selection, rate and price analysis, marketing, product development, fraud detection and prevention, as well as loss control.
Data users and suppliers no longer are confined to the insurance industry. Statisticians, engineers, scientists and other experts in advanced analytics, climatology, medicine, geospatial and transportation technologies and fraud detection are joining the fray.
But are data-management best practices developed in the 20th century — data stewardship, data and quality standards and data-management tools — still valid in the 21st century?
Data stewardship is rooted in the concept that the business area is the caretaker of the information it manages. However, as the breadth, depth and value of data increases, organizations need to take a broader, more encompassing view (data governance) in directing people, processes and information technology to create the consistent and proper handling of information. In this age, governance needs to go beyond the data itself and directly address how data is used. And governance must extend beyond the company’s walls and encompass data managed by third parties.
New best practice: Corporate governance over data and data usage
It no longer is enough to promote and develop data and data-quality standards. Data managers and data users must take an active role in getting data and data-quality standards adopted — internally and with trading partners. It is only then that the benefits associated with standards — consistency, increased efficiency and maximized utility — can be realized. While business needs and uses must continue to drive the standards process, consideration also must be given to other potential and yet-to-be-defined uses, products and analytics. Designing standards to encompass future needs presents a great challenge to the data manager. But more so, it can be a great opportunity for the organization.
New best practice: Adoption of data standards designed to meet potential future needs
A growing number of tools and techniques can be used to support data-management and data-quality best practices, including: metadata repositories, data dictionaries, data models, data and process flows, master data management (MDM), detailed specifications, ETL applications, data profiling, audits and controls, data and text mining and encryption. Those tools must be used to address new data needs, such as protecting data, promoting transparency and supporting data reuse.
Metadata, data dictionaries and MDM are key to integrating data from multiple sources. MDM and metadata can link reference data from those sources. Once linked, compatibility and mapping of data content can be tested and achieved by use of metadata and data dictionaries.
New best practice: Continue to look for tools that promote and foster good data management and data quality, and apply the tools to new and emerging data-management issues
While best practices may need refinement and updating, associated data-management guiding principles remain true:
- Data should be fit for the intended use.
- Data should flow from the underlying business processes.
- Data quality should be managed as close to the source as possible.
In addition, data not only should be fit for the intended use, but its fitness also must be documented in case of potential reuse.
As data and data usage continue to explode, senior management and data managers must be diligent and nimble — not only applying time-tested best practices but also looking for continuous improvement in response to evolving business needs and changing priorities.
Peter Marotta, AIDM, FIDM, is enterprise data administrator at ISO, leading ISO’s enterprise data management activities. He is a past president of the Insurance Data Management Association (IDMA).
Side Bar: Tips for 21st Century Data Managers
- Define and follow enterprise data strategies: These strategies must be in sync with and in support of business strategies and plans and evolve as business strategies evolve.
- Support the interoperability of data within the organization and with trading partners: Ensure that data elements, associated business rules and data-transmission specifications are well defined where possible, making use of industry standards.
- Know and vet your third-party data resources: If third-party data is critical to the success of your organization, hold it to the same standards as your internal data. Know the ultimate data source, access data-quality characteristics — especially freshness, completeness and accuracy — and require complete and thorough data-element documentation.
- Treat data as a critical corporate resource: Control access to your granular data resources. Too often companies make data available in more detail than needed. As regulatory requirements and contractual restrictions tighten, so must access.
- Develop and implement comprehensive and flexible data-quality measures: Each data source must have associated data-quality expectations — expectations that can vary based on use and reuse.
- Remember that data management applies to more than structured data sets. It also must be applied to emerging unstructured data resources, such as text and imagery.
- Require adherence to data-management best practices at the corporate level, as well as at the desktop level: Much data aggregation and analysis now are done at the desktop; however, data-management best practices are not often implemented at this level. Data documentation, data quality and data controls must be applied throughout the organization.
This article originally appeared in IASA’s Interpreter and is reprinted here with permission of the Insurance Accounting & Systems Association (www.iasa.org ).
|
|
