Why accurate rooftop geocodes and persistent unique IDs matter

Geocoding lies at the heart of many insurance carriers’ operations and is the foundation of understanding location-based risk. Its use facilitates risk assessment, program development, product development, and underwriting processes. However, the industry has recently noticed several problems stemming from inaccurate or outdated geocoding data. Similarly,  while parcel-level, street-level, and ZIP9 geocodes have been the accepted level of accuracy geocodes for insurance carriers, they’re now being surpassed in value by rooftop geocoding because of its ability to better keep up with and track change and provide exact accuracy, encouraging insurers to dig deeper for even more accurate solutions. Switching to rooftop geocoding is a smart move for insurers, especially when paired with persistent unique identifiers (PUIDs).

3 potential challenges with parcel centroid, street level, and ZIP-level geocoding

Insurance providers have utilized parcel, street level, and ZIP Code geocodes to pinpoint their insureds’—or those seeking to be insured—properties, and this has generally been an effective way to process data. Yet street-level and ZIP9 approaches to geocoding pose potential challenges for rural addresses, non-postal addresses, new builds, and lots over an acre. Their efficacy falls short in these cases. Inaccuracy is the source of these specific problems.

Inaccurate information

Geocoding information has no required update time and relies strictly on when a geocoding service provider will update the information provided to them. 

Old information can cause catastrophic errors when preparing risk assessments. Street names change over time. ZIP Codes change over time. Building complexes are knocked down and rebuilt. Brands rebrand and move to new locations. People in high-rise buildings move away and sometimes switch units within the same building. All of these changes are constant and unpredictable.

Additionally, think of the many homeowners who add mother-in-law suites or separate walk-in basements to their homes. Without the updated information concerning additional residents, new plumbing, etc., the insured’s policy would provide less coverage than is probably warranted, leaving the client and the provider in trouble. This would further lead to awkward and frustrating encounters when a claim surfaces, resulting in a costly legal battle. Businesses run the risk of having to void the policy altogether.

Broad information

Address intelligence providers often claim to be rooftop accurate, but our research has shown that this accuracy is parcel centroid or interpolated street level. Again, this isn’t a problem for the general population involved in many address databases. But, especially in insurance, broad information that does not provide the specific unit information can lead to an insufficient or overcompensated claim that costs companies greatly in both the short and long run. 

If a geocoding service provider claims to provide rooftop accuracy, shouldn’t that mean the exact rooftop? When relying on a false claim of rooftop accuracy, insurance, and reinsurance providers, as risk modelers, conduct their assessments based on information that isn’t accurate enough. While the pin may mark the building with which an address exists, a rooftop location should be more specific to the area of the building where that unit is located. The inability of the parcel centroid and ZIP9 capabilities to accurately identify and link secondary address information to a specific place further complicates matters as insurance agencies cannot pass accurate data from their CAT modeler to their system and run accurate risk models. 

For example, consider a CAT modeler assessing a property for catastrophic storm risk. Assume that the geocoding information provided is accurate at a parcel-centroid level. The geocode pin is located at the center point of a large property. However, suppose the building has been built towards the back of the lot. A change of 500 feet in accuracy could significantly change the elevation from where the parcel pin is located because the building is built on a hill. In that case, this inconsistency can cause a discrepancy in the policies the customer qualifies for regarding coverage options and paid premiums. The location of the building in the CAT modeler’s report would be inaccurate, not because they did a lousy job but because the information wasn’t specific enough.

Another example highlighting this broadness issue occurs when one considers addresses not recognized by the USPS as deliverable or, in other words, non-postal. For instance, a rancher in a small town may pick up his mail at the PO Box for the city closest to his farm. Still, if the USPS doesn’t list his address as a deliverable address and a risk modeler bases their assessment off of the PO Box address, the ranch is likely to be highly under or overinsured. 

Broad information also becomes problematic when you need information on a specific unit within a business or building complex. The geocoding pin at a parcel or ZIP9 level is sufficient until you need to find the information for a more specific address, like a unit inside that building’s footprint structure. Insurers and reinsurers may need to know more specific data points to assess risk accurately. For example, information regarding what floor a client is on may change the level of fire risk their home has (fire and heat rise, so if other units combust below, their risk is higher).

False positives

Geocoding providers don’t always include address verification in their products. This can be a breeding ground for false positives, where an entered address matches an incorrect or non-existent address without an error message or sufficient indication that the address is incomplete or wrong. Address verification should be embedded into the geocoding process to prohibit false positives from entering your databases. Without it, agencies become much more liable for incorrect information.

Is there a solution?

Introducing PUIDs

Persistent Unique Identifiers (PUIDs) are the linchpin ensuring that address data remains consistent across disparate datasets and systems. Similarly, PUIDs embody persistence over time. Broken down, it’s essentially a unique key linked to a specific address. When connected with rooftop-accurate geocoding, a PUID consistently returns exact location data on any building or piece of land regardless of ZIP Code changes, city annexations, renamed streets, building demolitions, company rebrandings, and more. PUIDs enable seamless integration of information related to a property to improve risk modeling procedures, fostering a holistic approach to risk assessment and management.

Utilizing a PUID in conjunction with rooftop geocodes is the solution to the puzzle, and there are many practical applications for their use, especially in the insurance industry.