What is property intelligence?

Property intelligence can mean a few different things depending on what industry you are talking about, but generally refers to data-driven insights about a particular piece of land and any structures on it. For example, property intelligence for insurance tends to be used in the context of understanding risk profiles based on a combination of environmental factors, building characteristics, and potential hazards nearby. In real estate, property intelligence could refer to how the local school district impacts home prices, or how suitable a parcel is for a new development project.

Regardless of specific industry context, property intelligence is increasingly used by public and private organizations alike to make informed decisions. These decisions are powered by in-depth data analysis to produce critical insights that would otherwise not be available. There are many different types of data that can fuel property intelligence, including:

• Property attributes (number of rooms, roof type, year built, etc.)
• Tax records (property value, sales history, etc.)
• Demographics (average income, age, educational attainment, etc. of people who live nearby)
• Points of interest (number of amenities in proximity to the property, like restaurants, parks, etc.)
• Mobility (how many people visit the property’s neighborhood on a daily, weekly, monthly, etc. basis)

…and the list goes on. Essentially, if the data can shed light on a piece of land or building, it can be leveraged for property intelligence.

Geospatial data & property intelligence

While the types of data that are often used to derive property intelligence are all different, they do have one thing in common: a geospatial component. Geospatial data is foundational to property intelligence as it provides location-based context about specific places that can be analyzed to determine individual risk profiles, site suitability, and more. 

Most property intelligence analysis starts with a single address string, which is then geocoded to produce latitude and longitude coordinates so the property can be correctly plotted on a map. From there, additional datasets can be layered onto the map and analyzed together to derive detailed information about a property’s overall value, suitability for a specific use, or hazard risk.

Property intelligence for P&C insurance

Leveraging this geospatial data for property intelligence is particularly important in the property and casualty (P&C) insurance industry, especially given the increasing climate risks both policyholders and carriers must navigate in order to remain resilient. To put it in perspective, the amount of national catastrophe (nat cat) events per year in the US has more than doubled in the past decade, with over $1B in property damage occurring in 2022 alone. 

The rapidly changing climate and its impacts on property risk have caused P&C insurance rates to skyrocket and led some insurers to altogether cease operations in the highest risk geographies. However, the most innovative carriers are turning to data-driven property intelligence to not only increase their portfolio’s resilience, but also help their policyholders remain informed about their own dynamically changing property risk.

4 property intelligence examples for P&C insurance

There are many ways property intelligence is being leveraged in insurance to adapt to evolving climate risks and streamline data analytics. At Ecopia AI (Ecopia), we work with 7 of the top 10 US P&C carriers, plus some of the top reinsurers around the world, so we know a thing or two about how data is enhancing efficiency and resiliency. 

Here are the top 4 ways we see P&C insurers applying property intelligence today.

Reinsurance

One way P&C carriers balance risk across their book of business is by establishing a reinsurance treaty with another insurer. What this essentially does is transfer the financial liability of claims within a certain dollar amount to the reinsurer, preventing the original carrier from assuming too much risk. As you can imagine, a lot of analytics goes into defining specific reinsurance strategies, and property intelligence plays a key role.

For example, Harford Mutual Insurance Group uses building footprint and rooftop-level geocodes to develop strategic groupings of commercial properties in their portfolio to optimize their reinsurance strategy. Groupings are developed using an algorithm based on proximity to nearby structures and property value, enabling Harford to effectively balance their risk without manually analyzing individual buildings. 

Master data management

Developing rich property intelligence requires a lot of data, which can be difficult to manage at scale. While data holds so much potential for the P&C industry, many carriers still struggle to accurately relate disparate sources together and derive truly unique insights. Master data management (MDM) refers to eliminating data redundancies and linking different datasets together for deeper, more efficient analysis, and is a critical component of property intelligence.

Tokio Marine North America Services (TMNAS) relies on geospatial MDM to power their replacement cost estimation and risk assessment workflows. By taking a building-based approach to geocoding, TMNAS is able to easily identify the number of structures associated with an address and parcel to better inform policy underwriting. Before switching their approach to geocoding, TMNAS was underpricing policies and encountering much higher replacement costs than expected; now, they are able to identify the precise number of buildings and square footage that should be factored into the property’s risk assessment.

Risk scoring

The entire P&C industry is based on risk: predicting it, modeling it, mitigating it, responding to it, and more. However, risk management is no small feat given the individual situations of specific properties and the dynamically changing climate. It also requires combining a large amount of data to factor in all possible hazards, such as wildfires, flooding, wind, erosion, and even risks brought on by nearby structures (think shopping plaza with restaurants located directly behind a residential neighborhood). To effectively understand how all of these characteristics relate to each other and impact a property’s vulnerability, many carriers are devising robust risk scoring workflows.

Risk scoring enables carriers to simplify their risk assessment without overlooking the important details of all potential factors. Models can ingest multiple data sources and analyze them in relation to specific properties to ultimately derive a risk score. For example, a property located on a hill in a coastal area will have a risk score that reflects how the increased elevation will decrease flood risk, but increase landslide risk. It’s important to note that high accuracy geocoding is critical for risk scoring, as analyzing the wrong location (even by a discrepancy of just a foot) will undermine the integrity of the scoring algorithm and any decisions made from it. Innovative carriers like Neptune Flood avoid miscalculating risk by leveraging a building-based geocoding engine as the foundation for all of their underwriting analytics.

Imagery analysis

Most people are familiar with the saying “a picture is worth a thousand words,” but P&C carriers might alter that slightly to say “an image is worth a thousand data points.” Geospatial imagery captured by satellites, planes, street view cars, and drones provides insurers with a real-world view of properties to better understand risk profiles. However, for images to be turned into actionable insights, they need to be carefully analyzed or mined for data. For example, looking at an image might help an insurer see that there are many trees on the property that could fall during a windstorm, but to actually calculate that wind risk, the carrier would need to quantify the number of trees, measure their height and distance from the structure itself, and factor in historical wind data. 

While imagery alone doesn’t provide all of the data needed to develop true property intelligence, it can form the basis for the in-depth analytics carriers need to create and maintain an up-to-date database of risk profiles. P&C insurers are increasingly leveraging fresh, high-resolution geospatial imagery as a basemap for their other foundational data. For example, geocoding data can be overlaid on imagery to determine the quality of property coordinates, as well as fill in any data gaps that might exist for roof type, additional structures, or neighboring buildings. 

Where to get property intelligence data

It’s clear that property intelligence is a game-changer for insurance carriers looking to derive deep insights that help them build resilience against increasing climate risks, streamline analysis, and improve their bottom line in a hyper-competitive market. But the secret to effective property intelligence lies in the foundational data used to produce these insights, so it’s important to select the right sources. 

Here’s a breakdown of common property intelligence data sources, complete with pros and cons so you can make an informed decision when choosing data to fuel your own analytics.

Open source property data

As data science and geospatial analysis become more widely used across industries, including insurance, more and more open source data is available for anyone to download and use. While this data democratization is great for getting a general sense of patterns or trends, not all open datasets should be considered authoritative sources with which to make important decisions.

Pros: It’s free!

Cons: Most open source datasets have infrequent or ad hoc update schedules, so can quickly become stale. Open source data is also often inaccurate, incomplete, and messy depending on how it was created, so it is best used for approximations or very high-level reporting. 

Manual digitization

Traditional data creation methods like manual digitization are still viable options for producing property intelligence data. However, it is not the most efficient option available given the huge effort required to create the data and keep it up-to-date. 

Pros: You have complete control over data creation and can ensure accuracy.

Cons: It’s extremely resource-intensive, requiring hours upon hours of tedious imagery tracing and constant updating to reflect real-world change. And don’t forget - human error can also impact results.

AI-based data creation

Advancements in AI are driving innovation in property intelligence data creation by eliminating the need for manual digitization, ultimately saving time and resources for what really matters - the analysis. Keep in mind though, the accuracy of results produced by AI can vary, so be sure to thoroughly evaluate a source for completeness, accuracy, and freshness before using the data for any mission-critical analysis. 

Pros: It’s accurate, efficient, and cost-effective. 

Cons: Not all AI is created equal. Some AI digitization tools or data providers can produce results quickly, but sacrifice some accuracy to do so. It’s important to take an AI-based, human-verified approach to avoid losing data quality. 

Get started with the highest quality AI-based property intelligence data

Ecopia has been in the AI space for over a decade, and our machine-generated, human-verified data is currently powering complex property intelligence analytics around the world. Our AI-based systems created the first and only complete map of buildings in the US, which we keep updated based on the freshest geospatial imagery available to accurately reflect real-world change. We use this foundational building footprint data to fuel our Building-Based Geocoding engine, complete with a robust system of unique identifiers to facilitate better master data management and high-precision risk assessment for our insurance customers. 

To learn more about AI-driven property intelligence, get in touch with Ecopia’s insurance team.