Seattle's Two-Speed Office Market: Why One Tenant-Rep Playbook No Longer Works in 2026

Most "state of the market" reads this spring will tell you Seattle's office market is recovering, and they have a number to prove it. Puget Sound recorded +253,352 sq ft of positive net absorption in Q1 2026, the first positive quarter since Q1 2022, with vacancy and availability each falling roughly 90 basis points, per Newmark's Q1 2026 Seattle/Puget Sound Office Report.

That single regional number is the most misleading figure in the market right now. It's an average, and there is no average tenant. There is no single "Seattle office market" in 2026 to recover. There are two, and they are moving in opposite directions at once.

Here is the split, in the numbers that actually decide a tenant-rep engagement.

Downtown Seattle is a tenant's market at a level the region has arguably never seen. Downtown CBD vacancy reached 36.5% in Q1 2026, up roughly 350 basis points year over year (Cushman & Wakefield Q1 2026 MarketBeat). On Cushman's national table, Seattle's metro vacancy of about 33% is the highest of any major U.S. market, against a national average near 20%. Asking rents are soft and getting softer: Seattle CBD Class A slipped to $51.01/sq ft, down $1.34 year over year (Newmark Q1 2026). Lee & Associates pegs broader Seattle vacancy at 16.79% with average asking around $37.30/sq ft, a reminder that the 36.5% is a downtown-core figure rather than the whole city.

The Eastside is the opposite story. Bellevue CBD Class A asking rents rose to $67.58/sq ft, up $6.54 year over year (Newmark), and downtown Bellevue now commands roughly 45% more than the broader Seattle metro (CoStar). The driver is an AI and tech land grab. OpenAI expanded to more than 272,000 sq ft at City Center Plaza, more than quadrupling its footprint (CoStar), with Uber, Meta, Statsig, Anduril, and xAI among the other recent Eastside signings (Lee & Associates; Marcus & Millichap). No new supply is coming to relieve it: per Newmark, there are zero office projects under construction region-wide, a first for the region.

Same metro, same quarter, a Class A rent gap of about $16.57 between the two cores, with the rent lines pointing in opposite directions.

An honest read can't stop at the bullish data, because the demand side carries a warning the headline absorption number hides. Seattle metro unemployment climbed to 5.4% in February 2026, up from 4.3% a year earlier and above the national rate of 4.4% (Newmark). WARN-notice layoffs hit 3,420 in Q1 2026, up 56% from Q4 2025, dominated by Amazon's 2,387 cuts, with additional reductions at Meta, Oracle, Atlassian, and Epic Games.

So the recovery is real but fragile, and it is concentrated. AI and advanced-technology users are doing the absorbing while traditional knowledge-economy hiring is still shedding seats. For a tenant-rep broker that nuance matters: the demand isn't broad-based. It comes from a specific tenant type, in a specific submarket, on a specific clock. That is why one playbook can't cover the book.

For fifteen years, the tenant-rep edge in Seattle was knowing the market: which landlords were soft, which subleases were quietly available, what a fair concession package looked like. That knowledge was the moat.

A two-speed market quietly retires that moat. Market knowledge still matters, but the right answer is now opposite depending on which side of Lake Washington your client sits on, and the window to act is closing at different speeds on each side.

Run the downtown playbook on an Eastside client and you'll counsel patience into a tightening market with no new construction and a shrinking sublease pool. Newmark reports Puget Sound sublease availability has fallen to 3.1%, down from a 4.3% peak. Your "let's wait for a better deal" advice ages into a rent increase. Newmark's Pacific Northwest research director, Yemas Ly, names the dynamic directly:

Zoom out and the same pattern repeats. CoStar's Phil Mobley notes U.S. office tenants signed roughly 120 million sq ft in Q1 2026, the strongest quarter since 2018, but it was "driven by an exceptionally large number of transactions rather than a resurgence of large deals," with lease sizes still about 15% below pre-pandemic averages and a "lock-in effect" keeping large, well-leased Class A blocks scarce.

Translated to Seattle, the volume sits in mid-sized requirements: the tenants Yemas Ly says are about to feel the Eastside window close, and the tenants who still hold the whip hand downtown. More, smaller deals, splitting in opposite directions. A broker running both sides on one instinct will be wrong half the time.

We built BrokerHQ on the thesis that the next decade of tenant-rep advantage comes from intelligence density rather than headcount: a live command center that combines tenant rosters, lease-maturity signals pulled from public filings, and real-time leasing activity in one place a broker actually works from. The 2026 Seattle market is the cleanest argument for that thesis we've seen.

When the market had one speed, you could win on relationships and a good memory. A two-speed market punishes the broker whose information lives in their head and their inbox, because the right move on Tuesday for a downtown client is the wrong move for the Eastside client they're also representing, and both windows are moving on different clocks. The brokers who compound through this cycle are the ones who let software hold the live state of the market so their judgment can move at the speed each side now demands.

That speed point has a deeper edge to it: it's less about doing the old job faster and more about showing up better prepared than the broker across the table. We made the full argument in a companion piece, "The AI Advantage in Tenant Rep Isn't Speed. It's Credibility."

I'd be a poor operator if I didn't name the other side. Three caveats worth holding.

First, the data carries real methodology noise. Regional vacancy is reported anywhere from about 16.8% (Lee & Associates, broad market) to 23.2% (Newmark) to 33% (Cushman & Wakefield's major-market table) to 36.5% (C&W downtown CBD), depending on geography, asset class, and whether sublease space is counted. The direction is solid and the bifurcation is unambiguous; any single decimal is not. Cite the firm and the measure every time you use one.

Second, the recovery is concentrated and the labor data is a genuine warning. A 5.4% metro unemployment rate with rising layoffs is not the backdrop of a durable boom. If AI-tenant demand cools, the Eastside thesis softens fast. Advise accordingly: the leverage-window narrative is a base case, not a certainty.

Third, "two-speed" is a frame, not a law. Submarkets blur at the edges. South Lake Union biotech, the Spring District, and suburban value plays all sit between the two poles. The frame is a starting diagnosis, not a substitute for pulling the actual comps for the actual client.

Stop reading Seattle as one market. In 2026 you are representing clients in two markets that happen to share a metro area, and the playbook inverts depending on which one your client sits in. Downtown rewards patience and leverage; a 36.5% vacancy core still favors the tenant who waits and negotiates. The Eastside rewards urgency and optionality; zero new construction, a shrinking sublease pool, and AI demand mean the favorable-terms window is closing on mid-sized requirements first. The thing that lets you run both correctly, for different clients, in the same week without your information going stale is no longer a sharper memory. It's a faster operating system.

The consensus call from CBRE, Newmark, Kidder Mathews, and Lee & Associates is gradual improvement through H2 2026, with the Eastside leading and downtown lagging. That consensus will be right in aggregate and useless in practice, because "the market improved" will again hide two opposite stories. The brokers who win the next twelve months are the ones who stop waiting for the market to turn and start running each speed on its own clock.