BuildingFootprintUSA delivers high-quality location data to help insurance companies make accurate risk assessments and pricing decisions. VP of Strategy Scott Robinson explains how they are creating new products and models to meet location intelligence challenges.
Can you tell us about your background and your role at BuildingFootprintUSA?
Scott: I’ve been in the mapping and Geographic Information System (GIS) space since 1992, including 15+ years in the MapInfo, Pitney Bowes ecosystem as a partner and employee. I’ve also spent a decent part of my career in technology and product management, mostly managing data products for companies like MapInfo and RMS.
As VP of Strategy at BuildingFootprintUSA, I spend a lot of time thinking about the what and the why. What is our product? What does it do and why does it matter for customers?
Who are your customers and what are they using your data for?
Scott: We work with insurance companies in the Property & Casualty space. Those could be carriers, reinsurers, catastrophe modelling companies, Lloyd’s of London, etc. We also have customers in wildfire underwriting in California and the flood certification space.
Typically, insurance companies are interested in the precise location of a structure so they can ask questions of the data. What is the property value? What hazards are impacted by the location? If they know the precise location, they can tell if it’s in or out of a flood zone, or how far away it is from a hydrant or fire station.
Our data is used for pre-fill and enrichment, risk assessment and risk aggregation. Insurers do accumulations of their risk and knowing the right location makes those accumulations a lot more precise. We also have customers underwriting risk for complex structures like college campuses or hospitals. Being able to disaggregate those buildings from a single address into multiple structures is something our customers get a great benefit from.
What are the challenges around location data and how does BuildingFootprintUSA differentiate itself?
Scott: A lot of the time, data and individual data components are not delivered as a solution. Businesses are piecing together integrated solutions with low vendor support, and we try to address that by pulling everything together and doing the heavy lifting.
We also see the market waiting for data components that haven’t been invented. They need to look at a structure and ask questions about it. Things like construction type, the year it was built, or the number of stories. There are solutions out there trying to provide that, but they aren’t complete enough to provide confidence for an underwriting team to make decisions.
A lot of data components are commodities, and a lot of companies in this space don’t try to differentiate on that data. They resign themselves to using imperfect data, and what we’re trying to say is there is something better out there. There are ways to differentiate the data through a fresh approach.
How are you finding those precise locations and where does the data come from?
Scott: We build accurate, enriched data solutions at the building level to help our customers, using what we call a curation model. We work with open data and in the United States, a lot of that comes from the county or municipality level. They have GIS expertise and build data sets to allow them to be able to collect taxes.
We believe that this data is going to be much better than anything derived from machine learning, imagery or LiDAR sources. We’ve built up that expertise over many years through relationships with over 1000 different government organisations.
What about third-party data sources? How do they fit into what you’re doing?
Scott: While that first piece of the curation model is really important, it doesn’t make it complete. To fill in gaps where we don’t have open data, we extract from either imagery or LiDAR to build out structure polygons. We can then plug those into our manufacturing capabilities to add in enrichment and attribution.
We’re always looking to bring in complete third party information that we can integrate at the building level because that typically hasn’t happened before. It’s typically been aggregated to the higher level and we’re attempting to bring it down to the structure level. Adding things like hazard data, building permits, and assessor information makes for a very innovative product.
We’re pushing the envelope with assessor data. It’s been available for many years, but when insurance companies want to use it, the fill rates (available information) don’t meet expectations. Variables like age and construction type are critical for assessing risk, so we’re creating new products and models to get those fill rates to a much higher percentage.
How do you deliver the data to customers?
Scott: We provide our data via both delivery and popular mapping and GIS formats, and we’ve just launched a new API. This allows the customer to integrate results into their geographic information system, underwriting platform, or claim systems, etc.
Customers can have thousands of locations that need enriching with key information for making accurate risk assessments and pricing decisions. The API allows us to deliver data in a transactional fashion and apply some of our expertise around geospatial capabilities. For example, we can find the nearest buildings within a certain area that meet specific characteristics.
Congratulations on the launch of the API. Is that a big focus for the company in 2021?
Scott: It’s the most important activity for us and we’re stepping up the launch in the first quarter of this year. We spend a lot of time talking to insurance customers about how best to integrate data into their workflow. Demand for an API is something we’ve heard over and over, especially from customers in the Lloyd’s market.
It was the number one thing in terms of product enhancements or future direction and delivering that changes our business.
What else is on your radar in 2021?
Scott: We’ve always been focused on expanding our coverage. Two years ago, it was somewhere between 55% and 60% of the US population. Today we sit at about 85% and we’ll get to 95% in 2021.
Also, as I mentioned earlier, customers love assessor data, but the fill rate just isn’t good enough. We’ve put a heavy emphasis on modelling out a concept that we call Assessor ++. We’re going to start with construction type, year-built numbers, number of stories and square footage. It’s something we’re extremely excited to bring to market this year.
Once we’re able to prove out further, we can go into different areas using machine learning and AI to model different attributes that are necessary or critical for property underwriting. We’re interested in assigning an address to every structure on the most granular level possible.
Where do you see the big opportunities coming from in your space? What can we expect to come next?
Scott: I would say around machine learning and artificial intelligence at the structure level. There are companies out there doing that, but it’s still at a kind of phase one. Seeing how that evolves and how companies engage and connect into this notion of property level machine learning and AI will be interesting.
We’ve also heard a lot of companies talking about automated data-driven inspections. Covid-19 has had a hand in that and, companies tell us the whole notion of inspections will become a lot more automated. I see significant disruption in those areas in the next few years.
We’re delighted to have BuildingFootprintUSA as a member of InsTech London. Why did you join and who are you looking to connect with?
Scott: We view InsTech London as the premier insurance technology community for the UK market. You have outstanding reach, especially into Lloyd’s, so there’s no better source for reaching that market than InsTech.
We’d like to connect with P&C carriers and reinsurers who are looking to improve their underwriting capabilities with better building level data. We’d also like to connect with insurtechs and channel and ecosystem partners who are interested in exploring the integration of building level data into existing or potential offerings.
Our building level analytics are different from what most insurers have been using to inform property underwriting. Get in touch and we can show how we think that happens and how it can impact a business.
For more information on BuildingFootprintUSA, go to buildingfootprintusa.com