At a packed conference booth, Annette Cotton and Matt Key were answering a question they’ve heard constantly from GIS professionals and real estate firms: Can I get it in ArcGIS?
Cotton is Chief Data Officer for First American Data and Analytics, with four decades in the data industry. Key is VP of Property Data and a 20-year company veteran. Together, they represent a company at an inflection point. One is sitting on one of the most comprehensive property data ecosystems in the country and is finally making it natively accessible within the tools that spatial analysts actually use.
A Long Time Coming
First American has been an Esri partner for five years, but the relationship was always incomplete. Their data was available through bulk delivery, APIs, and their SaaS platform, just not in the way Esri’s users wanted it.
“People would say, ‘Hey, that’s wonderful, but can I get it in ArcGIS? Can I get it as a feature service?’ “Key recalled. The answer was always a reluctant workaround.
This year, that changed. First American now delivers its data via feature service directly inside ArcGIS, opening what Key calls a brand-new market of spatial professionals who need property intelligence embedded in their maps, not bolted on from the outside.
The Problem No One Solved: 3,143 Counties
American property data is fundamentally fragmented. The country’s 3,143 counties never agreed on a standard way to collect or share property information, producing a patchwork of schemas, identifiers, and document formats that can vary even within the same county.
What First American has spent decades doing is collecting that data sometimes online, sometimes in person, sometimes via a CD mailed from a county clerk, and standardizing it around a common property identifier that ties together a parcel’s location, ownership history, transactions, and valuation regardless of jurisdiction.
That identifier does more than organize data. When Phoenix began renaming numeric streets to named ones, properties that had always carried one address suddenly had new ones in county systems. Without a managed identifier, the historical record breaks. “The IDs just preserve the links and connections with that change so that you’re not disrupted,” Cotton explained.
For GIS teams, the payoff is simple: click on any property, anywhere in the country, and get a consistent, reliable answer.
From Data to Decisions
Once that foundation is in place, the use cases multiply. First American is building an AI-powered app on Esri’s platform that lets users ask natural-language questions of the property data, something like: Tell me every vacant parcel in this ZIP code owned by someone who lives out of state.
Vacant land is the hottest area right now. Builders track competitor acquisitions to target different markets. Investors priced out of single-family homes by high mortgage rates are turning to vacant parcels as an alternative asset class. Cities hunting for housing infill opportunities can map every parking lot in their jurisdiction alongside ownership, value, and zoning data, turning a policy problem into a spatial one.
First American also produces analytics on top of raw data: Home Price Indices, property valuations, and a new Automated Valuation Model for vacant land, which has historically been nearly impossible to price without comparable sales.
The Trust Problem in AI
Ask what comes next, and the conversation turns to AI, but with more nuance than typical conference enthusiasm.
A recent survey of 350 commercial real estate professionals found that people liked AI but didn’t trust it enough for real business decisions. Key’s read: the problem isn’t the AI; it’s the data underneath it. “If you have quality inputs, your outputs should also be quality.”
This is where First American’s position as an actual manufacturer of property data matters. There are perhaps three or four companies nationally that collect and produce property data from sources. First American is one of them, and because its parent company uses that same data to underwrite title insurance, data quality isn’t a marketing claim. It’s existential. “Everyone can say quality. It’s an easy buzzword. For us, we have to live it every day.”
The team is also quietly developing something striking: using AI to scan archived document images and identify who signed on behalf of LLCs that own property. An estimated 40–50 million U.S. properties are held by corporations or LLCs. Finding the human behind “Gold Rush Properties” used to mean digging through county archives. Now it’s becoming a query.
“We have conversations every hour where someone says, ‘I didn’t know you could do that,'” Cotton said.
A New Chapter
For a company most people associate with title insurance, First American Data and Analytics is telling a different story, about the depth of property intelligence it has spent decades building, and what becomes possible when that data lives inside the tools where spatial professionals already work.
“I want us to be earning the reputation that we deserve,” Cotton said. “Highest quality data.”
