Insurance is an extremely competitive industry. To stay afloat and keep up with the current market, insurers must improve the accuracy of their risk models while keeping their pricing affordable. But that can be difficult when not every benefit can be measured the same way, and policies have to be written on a case-by-case basis.
As a result, many insurance firms are turning to big data to get faster and more accurate insights from a wider variety of angles. But what specific kinds of data do insurers need, and where can they get it?
We at SafeGraph recommend using location data in the application of data science in the insurance industry. We’ll expand upon how POI, property, and foot traffic data are essential for building precise liability policies by looking at how insurance providers can use geospatial data. We’ll also suggest a number of great places to get the location data you need to supercharge your insurance operations. Here’s a brief rundown of what we’ll be discussing:
Let’s start by looking at some of the services insurers provide that can benefit from applications of big data in the insurance sector.
Whether insuring a property, an organization, or a person, insurance firms perform many different functions. And to perform them well, they need the right types and quantity of data to understand who their clients are and what their unique needs are. Here are some examples of insurance applications that big data is critical for.
Now that you’re familiar with a few big data applications in insurance, the question becomes: where can you get the kind of data that your company needs? We’ll make your hunt a little easier by providing some recommendations in the next section.
As we mentioned, different types of insurers need different types of data. For example, a property management company requires a more thorough analysis of the environment and building criteria, whereas an auto insurance company would simply make their decisions based on urban hazards and the driver’s personal record.
Here, we’ve listed seven data providers that we recommend buying location data from, at least one of which will hopefully fit with your insurance niche.
Cost: Contact for pricing
Best for: accurate geospatial information
SafeGraph is a leading provider of US, UK, and Canadian location data. Our Places and Geometry data sets provide comprehensive attribute and spatial modeling information (respectively) for over 9 million points of interest across the three countries. Our Patterns and Weekly Patterns data sets also provide visit counts to over 4.5 million points of interest in the US and Canada.
These data sets provide insurers with a wealth of information that can be used to write more accurate policies. This includes the location and layout of a business; what environment or other facilities surround it; how many people typically visit it; the times, dates, or seasons when it's busiest; how long people usually stay there; and more.
Cost: $8,000-$10,000/month; custom pricing for enterprise
Best for: consumer identity management
Infutor has property data like SafeGraph does, but having a huge breadth of consumer demographic data is their main focus. They draw from an anonymized mix of communications information, retail transaction histories, census-based demographic profiles, automotive ownership records, and more to paint a picture of what Americans are buying and why.
This is valuable information for brokers selling personal life insurance, as it allows them to craft a policy based on the amount of risk associated with people’s typical lifestyles in a geographic area. It also helps automotive insurers by tracking the conditions of cars bought and sold in an area, giving an indication of how safely people drive there.
Cost: $0.05/record, charged on a per-dataset basis
Best for: climate-based indicators for general neighborhood data
ClimateCheck is a great resource for home and life insurance data. It combines information on over 140 million US properties with historical weather and climate data, processed through an aggregation of over 25 internationally-recognized climate change models. This results in an easy way to assess environmental risk to a property (and the people who live there) from fires, heat waves, storms, droughts, floods, and more.
Cost: $0.05/record, charged on a per-dataset basis
Best for: property values and title assessments
First American provides many valuable data sets and resources (including Calculators) regarding property attributes. These include a building’s location, zoning, and other characteristics, but also include a lot of useful-for-insurance financial data as well. For example, First American’s data tells you how much properties have sold for, if they’ve been mortgaged (and for how much and for how long), what they’re valued at now, and if they’ve been financed (and by who and for how long).
Brokers of homeowner and other property insurance can take this data into account when writing policies. It can tell them not only how much physical risk tenants will take on, but also the degree of financial risk tenants will face when comparing their assets against the value and transaction history of the property.
Cost: $0.005/record; custom pricing for enterprise
Best for: assessing information for various categories of property
BA45 has data on over 125 million commercial, residential, and government properties across the US. This data is segmented into over 40 attributes, including the type of building, the geography of the land it sits on, when it was sold and for how much, who currently owns it, and more. That means there’s lots of information there for insurers to assess both physical and financial risk for a property.
Cost: $0.05 per record; can vary from $7,000 to $125,000/year based on level of detail and whether purchasing for the entire US or a single state
Best for: spatial data for varying geographic areas of the US
Regrid offers lots of amazing spatial data options. Their data sets cover over 150 million parcels of land in the US, accounting for over 99% of inhabited land in the country. These data sets can be purchased by county, by state, or for the whole country. Each parcel can come with up to 120 different attributes, including a standardized address, land use classification code, value, tax bracket, geometric boundaries, and more. They also provide information about buildings on a property, including their characteristics and footprints.
This is another great company to buy geospatial insurance data from if you’re looking to insure people or places in the US. Since Regrid allows you to narrow down data sets to individual counties, you can buy only what you need without overspending and incorporating excess data that could muddle your analysis.
Best for: demographic data collected by the US government
The United States Census Bureau shares their data from surveys and censuses publicly on their website, making statistics, graphs, maps, and other information available for free. This includes data on standard attributes such as age, education level, income, gender, and so on. But in a specifically American context, it can also provide life insurance survey data. This can include uninsured versus insured participants (and their demographics), private versus public healthcare data, popular health insurance types and plans, etc..
None of this is precise location data, as the US government purposely removes individual identifying information for privacy reasons. However, if combined with any of the data sets above, it can allow insurers to apply big data to life insurance for writing more personalized plans at more accurate price points.
Using data science for insurance allows firms to better assess and predict risk for people and properties alike. The more accurate this analysis, the better insurers can underwrite policies at prices that stay competitive in a market that is becoming increasingly unpredictable.
A great example of this is insurance risk modeling with geospatial data. This combines point of interest data, foot traffic data, and environmental data to create a comprehensive picture of potential insurance risks to specific locations.
In short, various types of data will be important to the insurance sector for a long time to come. So it’s critical to capitalize on this trend to improve the accuracy of your policies while still keeping them at competitive prices. SafeGraph has more on how to precisely assess insurance risk with geospatial data, and you can visit our data shop to test-drive some of this data for yourself.