I spent the past two days at CREtech New York walking the floor, sitting in on panels and taking as many startup demos as I could. AI was everywhere, as promised. The headline sponsor's banner asked "Who's accountable when AI gets it wrong?" before I'd even found my first coffee.
That question followed me around the big-company side of the show. The large owners, operators and advisers talked about AI mostly as a management problem. They want to know who owns the data, who signs off on a model's recommendation, and what happens when a leasing tool starts making pricing calls nobody approved. Session after session came back to governance, trust and workflows. I think that's a healthy sign. It sounds like an industry that has run a few pilots and wants the next round to hold up in front of a board or a lender.
The Innovation Stage felt like a different conference.
Nearly every early-stage company I talked to was building some version of GIS or data normalization. They're pulling together rezonings, site plans, permit filings, zoning codes and parcel data, which every city and county publishes in its own format (when it publishes them at all), and making it comparable across markets. Companies like CityScout, which surfaces development activity from public filings before a project is announced and Zoneomics, which maps zoning across thousands of jurisdictions so a developer can find out what's buildable on a parcel without reading a municipal code at midnight stood out.
Normalizing location data is the unglamorous core of this business. Site selection consultants and economic developers have done it by hand for decades, calling planning offices and reconciling one county's definitions with the next. They run into the same problem every week building the Location Intelligence tracker, since no two project announcements report capex, jobs or timelines the same way.
The data center sessions on the main stage kept returning to power. Developers can find the capital, but they can't always find the electricity, so they're rethinking where they build. When the grid decides the map, getting location data right matters a lot more.
So I came away seeing two bets. The incumbents are betting AI pays off once their own data is in order. The startups are betting the value is in knowing where, faster and cleaner than anyone else. For the people who read Location Decision, the second bet is the one to watch. The companies fixing location data are building tools that site selectors, EDOs and corporate real estate teams will be buying in a few years, and I wouldn't be surprised if some of the big names at CREtech end up acquiring them.
I'll be following up with some of these founders and executives in the coming weeks. If you were at CREtech and saw it differently, I'd like to hear from you.