GIS: High-Value Technology for Insurers

By Frank J. Coyne

In many ways, geographic information systems (GIS) technology is ubiquitous today. It's in the TomTom and Garmin navigation systems in our cars, embedded in iPhone apps, and easily accessible on the Internet through sites like MapQuest, Google Maps, and Microsoft's Bing Maps. We enjoy real-time, voice-command driving directions to any destination, and we can find the location of virtually any commercial business, from the nearest Starbucks to stores accepting UPS packages on Sundays. A number of websites offer close-up aerial photography of our homes and neighborhoods. Geopositional systems in cell phones are saving lives in emergency situations, and in-vehicle devices such as GM's OnStar help police locate and secure stolen vehicles or contact emergency responders with the location of a car accident. The value of these commonplace GIS applications is readily apparent, and in a real sense they have already changed our lives.

Assessing the business value of any technology should be a straightforward endeavor approached with basic business questions in mind: Will it boost our productivity and ­profit­ability? Will it help us become increasingly service-oriented and competitive tomorrow? What can we expect in return for our investment? Fortunately, GIS technology has been held to high business standards for more than 15 years in our industry and has repeatedly demonstrated strong returns.

For its pricing and profitability, perhaps no other industry is as dependent as ours on the specific location of its customers. Insurers rely heavily on location to make their most basic underwriting and rating decisions and to define rating territories for personal auto, homeowners, commercial property, and other lines. Insurers underwrite and price risks based on their inclusion within coastal beach plans or within regions prone to hurricane, earthquake, flood, or wildfire. Property underwriters need to assign risks to the appropriate classified area for public fire protection.

Insurers, reinsurers, and reinsurance intermediaries rely on risk concentration analysis for catastrophe assessment in treaty negotiations. And when disaster does strike, claims managers use GIS technology to generate detailed maps showing the location of the affected risks, enabling rapid response to assist their insureds.

It's hard to imagine doing any of that now without GIS. And the good news is there's more value to come from this burgeoning technology.

Consider auto rating. The typical rating plan factors the driver (credit, age, etc.), the type of car, and the territory of the garaged address. This is changing in two big ways, and GIS plays a pivotal role in both.

First, many insurers currently divide the country into about 1,000 territories. But thanks to GIS and predictive modeling, we're uncovering significant heterogeneity within established territories and can predict loss much more accurately using up to 200,000 micro-territories in the United States. One such approach is already filed in more than a dozen states.

In addition to territory refinement, GIS is at the center of a fascinating technology mix including GPS, cellular communications, and vehicle diagnostics — currently called vehicle telematics. This technology will help replace the surrogate variables insurers now use to predict driving behavior and loss with intelligent assessment of actual driving behaviors of insureds. We'll be able to determine behaviors such as speeding and repetitive hard braking, the frequency of driving in high-traffic-density areas, and the radius of operations. We'll also be able to ascertain true garaged address with great accuracy. The trend presented by this technology is called usage-based insurance. It's hard to remember — or even imagine — a greater opportunity to reduce the societal and insurance costs of ­auto­mobile accidents.

Next, consider the insurance impact of climate change and the frequency of diverse weather events. Understanding how weather affects society and property allows us to translate climate change projections into useful future risk scenarios. From the insurance perspective, it is critical to correlate the impact of climate changes and the probability distributions of some extreme weather events. Climate already affects the spread of wildfire-exposed risks, mold exposure, hurricanes, loss due to freezing, and water-damage losses. Climate and weather can also dramatically influence consumer demand, retail supply chains, and the financial markets. For example, a recently created weather index by ISO subsidiary Atmospheric and Environmental Research (AER) has explained approximately 85 percent of the variability in September sales results for Kohl's department stores.

Academic climate research is quickly turning into applied results. Increasingly, industries are asked to disclose their policies, procedures, and risk mitigation actions regarding climate change. In a recent National Association of Insurance Commissioners (NAIC) survey, seven out of eight questions were specifically directed at understanding how insurers' risk management strategies account for climate change. AER also participated in a recent NAIC Climate Change Summit, and its renowned research experts are improving techniques to couple remote sensing with GIS. Those techniques provide timely and actionable updates of land use and land cover reflecting changes due to fire, flood, clear-cutting, crop health, and other factors.

Risk managers in a number of industries are using GIS-based models to help plan loss control activities. They develop GIS crime indices to route delivery fleets safely, allocate security expenditures across company locations, and assess the ­likeli­hood of financial loss due to crime. And their GIS modelers are helping them estimate the financial impact of insured property and workers compensation losses from potential future terrorist attacks.

GIS data speaks strongly to insurers' marketers too. When rating and hazard variables are overlaid in a GIS with demographic variables, key sales and expansion opportunities can be uncovered. Insurers can use a GIS application to identify attractive niches. For example, it is fairly easy to run a direct-mail campaign targeting dwellings exceeding $300,000 in value, in established low-crime communities, with low to medium catastrophe exposure, and located in ISO's Public Protection Classification (PPCTM) tiers 1 through 7. Marketers have used demographic data extensively in the past but have not always integrated buyer characteristics with rating and hazard information at the address level of detail. This is changing.

GIS is available technology with immediate value for insurers. And ISO is excited to be a leading innovator in GIS applications for risk assessment, dedicated to developing new solutions to industry needs in the 21st century.

Frank J. Coyne is chairman, president, and chief executive officer of ISO.