What Seattle's WARN Notices Actually Tell a Tenant Broker

The WARN Act requires companies above a certain size to give 60 days' notice before a mass layoff or plant closing. That notice becomes a public filing with the state. In Washington, that means every WARN notice filed in the trailing 12 months (Aug 14, 2025 to Aug 14, 2026) is sitting in a public dataset, entity-keyed, timestamped, and free to query.

That's the whole appeal and the whole trap. A WARN filing is real. It's not a rumor, not a LinkedIn post, not a secondhand tip from a broker friend. A company told the state, in writing, that it's cutting a defined number of jobs. But 'cutting jobs' and 'giving back space' are two different facts, and WARN only gives you the first one.

34, as of August 14, 2026. That's the count of Washington WARN notices with a signal_date in the trailing 12 months ending that date. Not 34 companies, 34 notices, because a single company can file more than once (amended notice, phased layoff, a second round six months later).

Deduplicate by company and the number drops to 19 distinct entities. That's the real denominator if you're asking 'how many Seattle-area companies had a layoff event on file this year.' 19 is small. It's not nothing, but it's small enough that a broker treating every WARN filing as a hot lead is going to burn through the list in an afternoon and then have nothing left.

Yes, and this is the split that actually matters for a broker's read. Of the 34 notices, 9 were filed as a full closure and 25 as a layoff short of closure. That's roughly one in four filings signaling the company is shutting the location down entirely, versus three in four signaling a headcount reduction at a location that, per the filing itself, keeps operating.

A closure filing is a stronger prior that a tenant might eventually give back space. A layoff-short-of-closure filing tells you staffing changed. It does not tell you the company is subleasing its floor, renegotiating its lease, or doing anything at all with its real estate. Plenty of companies right-size headcount and sit in the exact same square footage for years.

2 of the notices on file have an effective date still in the future as of Aug 14, 2026. That's advance notice already filed, not yet in effect, which is exactly what the 60-day requirement is designed to produce. If you're tracking these as a broker, that 2 is your closest thing to a forward-looking indicator: an event that is scheduled but hasn't happened yet.

It's still just 2 filings. Useful to know about, not enough to build a pipeline on.

The full WA WARN record for these companies goes back to Oct 1, 2006, 110 notices total across two decades. That history is heavily front-loaded toward the present: 26 of those 110 notices, roughly a quarter of the entire 20-year record, were filed in calendar 2025 alone.

That concentration is worth sitting with. It could mean 2025 was a genuinely elevated year for layoff filings in Washington. It could also mean some of these 19 companies are repeat filers with a long WARN history who happened to file again recently. The dataset doesn't distinguish 'first-time filer having a bad year' from 'serial filer, business as usual' without going company by company, and this brief doesn't have that breakdown.

Sometimes, and this is the one place WARN data earns its keep. Of the 19 distinct companies with a trailing-12-month WARN notice, 17 also appear in company_activity_events under at least one other independent source: a building permit, a food-service inspection, or another canonical signal type entirely.

That's a measured overlap, not an assumption, and it's worth being precise about why it matters. Independent public-record sources rarely line up on the same entity by coincidence. A permit filing comes from a building department. A WARN notice comes from a state labor office. Neither system knows the other exists. When 17 out of 19 companies show up in both, that's corroboration, a second independent government record confirming the entity is real and active, not a signal that the layoff caused the permit or vice versa. Don't read causation into it. Read it as: this entity is legible across more than one public system, which makes it worth a closer look, not worth an assumption.

It feels that way, and it's the reason WARN data gets oversold. But the logic skips a step. A company that lays off 30% of a floor's headcount can sit on that lease for its full remaining term, sublease part of it, renegotiate at renewal, or do nothing differently for years, WARN doesn't distinguish between those outcomes because it isn't a real estate filing. It's a labor filing.

The stronger counter-argument: with only 19 distinct companies and 9 closures in a full trailing year across the entire state, WARN is too sparse to be a standalone Seattle CRE lead source. Treat it as one data point to check against a permit filing or a sublease listing, not a pipeline on its own. If you're cold-calling off a WARN list expecting space to move, you'll be right occasionally and wrong most of the time, and 25 of your 34 leads were layoffs at companies that, per their own filing, aren't closing anything.

BrokerHQ's view: WARN data is real, it's public, and it's exactly as narrow as it looks. 19 companies in a full trailing year is not a pipeline, it's a watchlist. What makes it worth checking at all is the cross-source overlap, 17 of those 19 companies show up somewhere else in the public record too, and that's the actual signal: not the layoff itself, but confirmation that an entity is active enough to leave more than one footprint. Use WARN as a filter to decide who's worth a second look, not as a reason to believe space is about to move. We'd rather tell a broker 'here are 19 names and here's exactly what each filing does and doesn't say' than sell a layoff count as a leasing forecast.