ISO Review Industry Roundtable

In this ISO Review GIS roundtable, executives from the ISO Family of Companies share their views on GIS return on investment, GIS data quality, the evolution of GIS as a tool for insurers, how to take advantage of GIS today, and what new horizons GIS will allow insurers to explore in the future.

Phil Hatfield, J.D., CPCU, is vice president of product development at ISO Innovative Analytics (IIA). IIA is dedicated to developing new analytic products and enhancing existing products through strategic alliances with ISO business units, subsidiaries, and client companies.	Roger Grenier, Ph.D., is a vice president at AIR Worldwide. A member of the ISO Family of Companies, AIR is the scientific leader and most respected provider of risk modeling software and consulting services. AIR founded the catastrophe modeling industry in 1987 and today models the risk from natural ­catas­trophes and terrorism in more than 50 countries.
Venkat Rachakonda is senior GIS product consultant in ISO's Risk Decision Services unit.	Pram Lipczyk is manager, Risk Decision Information, in ISO's Risk Decision Services unit.

How should insurers assess ROI with GIS? How should they approach the build vs. buy decision?

Phil Hatfield: The ultimate purpose of GIS is to enable better operational decision making. While GIS investment is relatively easy to quantify, measuring the benefits can be more challenging.

GIS solutions are fairly specialized and rarely part of an insurer's core competency. The typical ROI analysis must determine whether it makes sense to develop and maintain a custom GIS solution in-house or if the right strategic decision is to purchase a state-of-the-art solution. In many cases, the use of an outside GIS solution makes economic sense because of economies of scale, expertise, and start-up costs.

Roger Grenier: It's important to determine the role of GIS within the organization regarding business opportunities and how GIS complements existing processes. To assess the costs and benefits for buying or building, it's critical to understand the material and human resources as well as investments required to obtain practical GIS data. Without skilled staff to execute and manage projects, hardware and software resources will not be fully utilized. A phased GIS approach that focuses first on data and personnel, and later on tools and process, will often provide the best ROI.

Venkat Rachakonda: Investment in GIS technology is expensive for insurance companies. But long-term ROI is high because GIS serves the customer with accurate, up-to-date information and helps insurers make more accurate decisions, leading to customer retention.

GIS technology is used primarily to build analytical and decision-making functions. Because the insurance business is unique and complex, even off-the-shelf solutions may need to be customized to deliver ROI.

What's transpiring in your GIS labs today, and what's on tap for tomorrow?

Pram Lipczyk: Our customer satisfaction surveys make it clear that the breadth and accuracy of GIS data must remain our primary focus. Relying on census data or any other single source of data for geocoding purposes is no longer an option. We've depended on multiple sources of address information for years, and we're constantly searching for new ones.

Rooftop-level address geocoding accuracy, however, is only part of the equation. Every spatial data set requires a comparable level of accuracy. It makes little sense, for example, to have rooftop-level accuracy for your policy addresses and only estimated locations for responding fire stations.

That's why we recently conducted field verification of all fire stations in most states, thereby ensuring the best possible geospatial accuracy of our data. We expect to execute similar projects in the near future.

Rachakonda: We also recently released LOCATION® Analyst to customers. The product was built on ISO's GIS platform landscape. Insurance agents and underwriters can use this tool to study their books of business and make precise and effective decisions. Future releases of the product will include enhancements and add-ons that incorporate even more statistical data maintained by ISO. Other ISO GIS insurance products scheduled in 2010 include a diverse range of spatially enabled applications and data sets.

Grenier: We're using GIS technologies to understand and convey information related to exposure data quality and concentrations of exposure that feed directly into the catastrophe models our clients use to make business decisions. Our team is developing tools to enhance the information the models produce, including global modeling solutions that can leverage the increasing volume of detailed GIS data becoming available worldwide.

How important is data quality in GIS, and how do you ensure the quality of data?

Lipczyk: The importance of data quality has increased in the last few years. Companies realize that to stay competitive they need more accurate and more refined data. Using publicly available hazard information — with infrequent or no updates — may have been sufficient a few years ago, but it is much less acceptable today.

Moreover, GIS users want to know if and how the underlying geodata has been collected and maintained. Is all the information collected over the phone and machine-geocoded? Is it captured from aerial or satellite imagery? Or is it validated in the field? Because of the evolving parameters, ISO deploys 600 professionals to field-verify many of its GIS data elements.

Rachakonda: When GIS data is of good quality, accurate, and up to date, the resultant decision making will be more effective. Nevertheless, improving GIS data quality increases costs and needs to be justified.

Five components control GIS data quality:

• lineage (collection and compilation methods, sourcing, ­geographic coverage, usage of spatial algorithms)
• spatial or positional accuracy
• attribute accuracy
• logical consistency
• completeness

Hatfield: As in all areas of analytics, data quality is critical in GIS analytics. Good analysis cannot make up for poor-quality data. Often the key data-quality criterion for analytics is simply good metadata. The analyst must understand how the data was collected and measured and how it was aggregated and ­inter­polated. That helps determine whether the data is fit for a particular analytic purpose and how GIS data from different sources can be combined.

Grenier: Catastrophe models have become critical to a wide array of business decisions. Recent experience with Hurricane Ike, for example, has underscored the importance that exposure data quality has on model results. GIS analytics is becoming essential for assessing the completeness and accuracy of the underlying exposure data.

GIS technology started life as a marketing tool. How is it evolving to help marketers in insurance make better decisions?

Rachakonda: The visual impressions created by more precise maps of customers, competitors, territories, and potential customers help insurers understand their marketing situations more clearly. Efforts to extend GIS to target marketing in the insurance sector are increasing and already realizing profits for some companies.

Using LOCATION Analyst, insurers can map their books of business and examine them demographically against various risks like crime, PPCTM, and wildfire, which might be concealed or overlooked in non-GIS analysis.

Hatfield: GIS can impact marketing decisions in three areas. First, through robust risk profiles of small geographies, GIS can help marketers target risks that fit their business profile. Factors other than location are required to underwrite a risk, but there is enough predictive power in geography to make marketing much more precise.

Second, micro-geographies offer useful information about the probable life-stage and psychographic profile of a prospect. A ZIP code may be too large and diverse an area to indicate whether a particular prospect might be an empty nester, but a life-stage assumption like empty nester is more likely to be correct for an individual based at a smaller geographic level like a block group. That allows marketers to reach and service intended audiences more accurately.

Third, GIS is critical for analyzing distribution. The majority of insurance is still sold through agency channels. In that sense, insurance remains a local business with customers seeking out agents located close to where they live or work. Therefore, it's important to have an agency footprint that matches the insurer's target customers. GIS enables that type of analysis.

What bold new questions will GIS allow insurers to explore in the future?

Lipczyk: One exciting aspect of the GIS industry is its constant progress and virtually endless opportunities. From the insurer's perspective, I see two opposite — yet in some sense complementary — trends. One focuses on an individual property and all questions that can and should be asked to assess the risk correctly. Some of the questions have been standard for years, such as the distance to a hydrant or fire station. And there are many new ones: How close is the nearest power line or cell tower? Do any neighbors have dogs, pools, or trampolines? While the GIS software has been capable of answering such questions for a long time, the data still has to be developed — and maintained.

The opposite trend looks at the entire policy portfolio and claims history. By analyzing the characteristics, including geospatial components, of a large number of policies that meet certain criteria, insurers will be able to define with extreme precision the detailed characteristics, or patterns, of risks that are most desired. While it sounds futuristic, GIS products that can provide such analysis exist today.

Hatfield: Better location information about risks and the factors that affect those risks will continue to compress the geographic component of risk analysis. Currently, typical territories cover relatively large areas — combining several counties or ZIP codes. But there can be a great deal of variability within a territory. Finer segments of the geography enable much more accurate risk assessment. The ultimate goal is to drive the analysis down to the specific geographic point where the risk is located.

And for risks that are movable, the addition of up-to-the-minute location information will enable the dynamic evaluation of risk in real time. This will also open new possibilities for customer service. We are already seeing trends like smartphone applications that allow a customer to report a claim soon after it happens — at the location of the accident and with pictures. This sort of innovative customer service will become more common as more customers use geolocation capabilities in their everyday lives.

Grenier: Leading companies are already using GIS technologies to assess the impact and mobilize response in the face of large-scale claim events, such as major wildfires, floods, and hurricanes. GIS analytics can also be used to identify growth opportunities — places where risks are not correlated with existing exposures and that represent areas for expansion. Managing exposure concentrations is central to a successful insurance enterprise and its goal to protect policyholders.

What should a good CIO do to ensure the company can take advantage of GIS today?

Hatfield: The presentation of information in the form of maps is a basic but underutilized capability. Maps are a powerful reporting paradigm to absorb information efficiently. Humans are hardwired to identify patterns in information that is presented graphically, and maps are a universal method for understanding geographic data. So, CIOs should ensure their organizations have sufficient capability to present location-enabled information effectively.

Grenier: Gather information from stakeholders to understand the business benefits that GIS can bring to the organization. Invest in developing or acquiring GIS expertise to understand the landscape of products and tools available that best fit the organization's goals. Finally, synthesize the needs of the organization with the available technical solutions to develop a coherent strategy that delivers the value GIS can provide.

Rachakonda: Traditionally, GIS was viewed as a niche ­tech­nology. But today, GIS has become one of the major IT platforms at the enterprise level in many companies. Companies that implement GIS have seen their business improve. Every technical department should realize the power of GIS and move forward to incorporate the technology. With proper planning and strategy, any organization can implement a successful GIS platform.

Lipczyk: Make sure different divisions or departments ­com­municate and collaborate with each other on GIS applications. Forming internal GIS user groups is a good practice. This may not only reduce external GIS expenses but also result in new opportunities. In many cases, one licensed GIS data set acquired for a narrowly defined purpose can be used successfully by other groups for a variety of purposes.