A commercial tenant does not decide to leave a building overnight. Long before a broker or landlord hears about it, that decision already shows up somewhere in public records: a layoff notice filed with a state labor office, a funding round logged with the Securities and Exchange Commission, a lease clause quietly counting down to expiration. The information exists, and it is public. The missing piece is a system that reads six such sources together, on a regular schedule, instead of leaving each one to a manual check that rarely happens.
For commercial real estate professionals in Central Oregon and everywhere else, building that system is worth the effort. It does not require a research department. It requires deciding that six public sources are worth watching together, then building the modest system that does it.
Six kinds of public records already carry the signal
Six categories of public filings correlate with a commercial tenant’s decision to move, expand, or shut down: layoffs, funding rounds, lease expirations, new business registrations, mergers and acquisitions, and bankruptcy filings.
Each one, on its own, is weak evidence. A single funding round does not mean a company is about to lease more space. A funding round paired with a hiring surge and a lease expiring within the next year is a different story, and catching that combination before a competitor does is exactly the job a system should be doing every day.
The scale of just one of these six feeds makes the point. Layoff notices required under the WARN Act, aggregated nationally by a service called WARN Firehose, cover more than 86,000 filings and 14 million workers across all 50 states. Federal rules only require that notice from employers with 100 or more employees who lay off 50 or more people at a single site, so smaller local businesses rarely trigger a filing. For the larger tenants who occupy the buildings most commercial portfolios care about, the paperwork already exists and is already public. Somebody still has to pull it, every day, from six different places, and right now that somebody is rarely anyone.
The data is scattered across six separate systems
This signal sits unused mainly because it is scattered: WARN filings live with state labor departments, funding and merger filings live with the SEC, lease data lives with county assessors or a handful of paid providers, new business registrations live with each state’s Secretary of State, and bankruptcy records live with PACER, the federal court system, which still charges by the page and was never built for anyone tracking hundreds of companies at once.
That fragmentation is a solvable engineering problem. Software built for other industries has already solved a version of it. Business to business sales teams pay enterprise vendors tens of thousands of dollars a year, sometimes more than $300,000 according to a 2026 survey of intent data pricing, to combine hiring, funding, and technology signals into a single lead score. Commercial real estate has no real equivalent yet, mostly because the audience is smaller and the underlying signals look different. That absence is an opening for whoever decides to build the matching system first.
What building it actually takes
A complete version of a tenant prediction system takes real engineering work. It has to match the same company across records that list it under three different names, “Acme Logistics LLC” in a WARN filing, “Acme Logistics, Inc.” in a state registration, and just “Acme” in a funding announcement, so one company does not get counted as three separate leads. It also has to weight each signal by age, so a bankruptcy filing from three years ago carries less weight than a layoff notice from last week. A firm does not need the complete version to start.
A workable starting rule looks like this. Rank the six signal types by how strongly each one tends to predict a move, since a bankruptcy filing is a stronger signal than a single funding round on its own. Give full weight to anything from the past two weeks, let signals older than a month fade into background noise, and treat two or three weak signals appearing together as more meaningful than any one alone. Check for a shared address, phone number, or company domain before merging two records into one lead. Whatever survives that filter is worth a phone call, and the firms that build this filter into a running system are the ones who will make that call first.
Some firms will build this with an in house developer. Others will work with a software team that already builds this kind of integration for property technology companies: combining data from a CRM, a real estate platform, and an AI model into one broker facing output is the kind of project Brocoders has delivered for a proptech client before. Who builds it matters less than recognizing that it is now a solved engineering problem.
Why the first phone call still matters more than the technology
None of this replaces the local relationships and market knowledge that already drive commercial leasing decisions. It changes the timing of when those relationships get used. Commissions on commercial leases commonly run from 3 to 6 percent of the total lease value, and a broker who spots a lease expiration paired with a funding round a month before a competitor does is the broker who gets that first conversation. The brokerages that build a system for that head start, even a rough one, will be the ones making the call. The rest will keep finding out after the sign goes up.
