The Oct 28 WARN Cliff: 8,220 Names to Pipeline Before LinkedIn

If you wait for the "Open to Work" ring to appear on a profile, you are already 60 days late. The WARN Act gives you a legally mandated forward calendar of exactly who will be looking for work and when, and the week of October 28, 2026 is the single densest cliff in the current window. Layoff Atlas is tracking 8,220 workers scheduled to separate that week, more than double any other week in the 90-day horizon.

Why Oct 28 is the week to pre-build against

The week of Oct 28, 2026 shows 8,220 scheduled separations on the Layoff Atlas forward calendar, 2.8x the average of the surrounding weeks. The follow-on cliff lands two weeks later on Nov 11 at 6,124. If you only build one pre-WARN candidate list this quarter, build it against those two dates.

Here is the full forward curve from the Layoff Atlas tracker, with the derived math a sourcer actually needs:

Nationally, Layoff Atlas counts 750 WARN notices covering 69,260 workers in the rolling 90-day window. The warnact.io YTD tracker shows 5,507 filings across 48 states and 586,710 workers affected in 2026. Pick one tracker, define the window, cite it. Mixing a 90-day count with a YTD count in the same memo is how you get torched in a hiring manager meeting.

8,220

What WARN actually gives you that LinkedIn does not

WARN is a 60-day legally guaranteed heads-up on separations, which means every name on an Oct 28 filing is still formally employed on Oct 20 and still unreachable on an Open to Work query. That 60-day gap is the entire alpha of pre-WARN sourcing.

The Department of Labor rule is simple: employers with 100 or more employees must provide at least 60 calendar days advance written notice of a plant closing or mass layoff affecting 50 or more employees at a single site. That is why the forward calendar is reliable rather than speculative. Boards do not file WARN to signal weakness; they file it to avoid treble damages.

Three things follow from the statute:

1. The window is 60 to 75 days of exclusive access. By the time Open to Work banners flip on, every competing sourcer sees the same inventory. The pre-WARN window is yours alone if you source by employer plus site, not by profile signal.
2. WARN is reported plans, not confirmed losses. Headcounts get amended, dates slip, some notices are withdrawn. Discount accordingly, but do not discount to zero.
3. WARN undercounts the real pool. Sub-50-person cuts, remote-worker layoffs with no "single site," and sub-100-employee employers never appear. The actual sourcing pool around any filing is 1.5x to 3x the WARN headcount once you count adjacent teams and survivor's-guilt attrition.

The 90-day map: Washington and California own 47%

Washington and California combine for 32,686 of the 69,260 workers in the current 90-day WARN pool, or 47.2% of the entire national total. Illinois (4,861), Texas (3,957), and Ohio (3,224) round out the top five. If you source US tech, those two coastal states are where the pre-build pays off most.

But geography is a trap if you stop there. Washington's #1 rank is not a tech story. Gebbers Farms of Brewster, WA filed a 3,445-worker agricultural notice that single-handedly accounts for roughly 21% of the entire Washington 90-day total. A sourcer who reads "Washington leads layoffs" and queues up ex-Amazon SDE pitches is misreading the data by an order of magnitude.

The right move is to filter by NAICS code or employer before you build any list. Named filings worth separating out of the raw feed right now:

• Gebbers Farms (Brewster, WA), 3,445 workers, agricultural. Not your pool.
• Spirit Airlines (FL), 2,529 workers. Pilots, flight attendants, ground ops. Different pool entirely.
• Tyson Foods (IL), 2,495 workers. Food processing.
• Bristol Myers Squibb (NJ), 4 notices totaling 845 workers. Public ticker, cross-reference the 10-Q.
• Del Monte Foods (CA), 3 closures, 4 notices, 822 workers. Illustrates that California's filing mix is heavily food and ag, not just tech.

The lesson: the headline-grabbing number is almost never the sourceable number. Read the employer column before the state column.

The 68 to 1 problem: why pre-build beats scrape-and-blast

Every engineer on an Oct 28 WARN notice is one needle in a 68-person haystack of indexed peers, so a successful pre-WARN candidate list is built by narrowing before the names are public, not by scraping after. In Refolk's index of professional profiles, there are roughly 560,792 US software engineers across Software Engineer, Senior, and Staff titles. Divide that by the 8,220 scheduled Oct 28 separations and you get a ratio of about 68 to 1.

That ratio is the reason "I will just search for recent ex-Company X employees on Nov 1" fails. By Nov 1 every other recruiter is running the same query, your inMails land tenth in a stack, and the strongest candidates have already answered a warm intro.

The pre-WARN playbook inverts that:

1. Pull the filing. Employer, site address, effective date, headcount.
2. Reverse the roster. In Refolk's Seattle-tagged query clusters, the dominant logos are Amazon, AWS, Microsoft, Expedia, T-Mobile, Google, and Apple. For a WA WARN at any of those, you already have the pre-cut roster indexed.
3. Narrow before the cut. Filter by team, tenure, tech stack. A 500-person site WARN is a 40-person shortlist if you narrow by "payments backend, 3+ years at the site."
4. Queue outreach, do not send it. Draft messages dated to land 7 to 10 days after the effective date, not before. Reaching out while someone is still employed is the fastest way to burn a lead.
5. Second-ring the survivors. The teammates who were not cut are the second cohort, triggered by survivor's-guilt attrition over the next 90 days.

This is the exact gap Refolk closes for pre-WARN work: you describe the person in plain English (senior backend engineers at Amazon's Seattle sites who have shipped payments or logistics systems) and get a ranked shortlist back across GitHub, LinkedIn, and the open web, not a 4,000-row CSV you have to triage by hand.

Reading the WARN forward calendar like a sourcing pipeline

Treat the WARN effective-date column as a 90-day pipeline with scheduled release dates, not a news feed. Each row is a dated candidate cohort with a known employer, a known site, a known separation window, and a legally bounded degree of certainty.

The mechanics behind the Oct 28 cliff itself are worth understanding before you brief a hiring manager. The spike is partly a calendar artifact: WARN's 60-day rule means late-August board decisions land in late October, and boards cluster approvals around quarter-end. Q3 earnings season is the forcing function. So "Oct 28 is heavier than other weeks" is true, and "layoffs are accelerating" may not be. Do not confuse one dense week with a trend line; the macro picture is bifurcated, with ADP showing private employers added 90,000 jobs in September driven by health care, education, and hospitality, while white-collar categories lost jobs.

A workable weekly cadence against the forward calendar:

• T minus 60 days: Filing appears. Log employer, site, effective date, headcount.
• T minus 45: Pull the indexed roster for that employer and site. Narrow to your shortlist.
• T minus 30: Draft the outreach, do not send. Verify emails and alternate channels.
• T minus 14: Monitor for amendments and withdrawals. Headcounts move.
• T plus 7 to 10: First outreach lands, warm, specific, acknowledging the situation without being ghoulish.
• T plus 30 to 90: Second-ring the survivors.

Trackers worth watching, and the ones that will confuse you

There is no single canonical WARN feed, so pick one tracker as your source of truth and only use the others for cross-reference. The federal DOL does not publish a unified real-time dataset; state agencies do, and third parties aggregate them with different windows and quality filters.

The ones worth bookmarking:

• Layoff Atlas (dev.to/layoffatlas): rolling 90-day forward calendar, best for the effective-date view used throughout this piece.
• warnact.io: YTD counts across 48 states, best for annual context (5,507 filings, 586,710 workers in 2026).
• LayoffAlert.org: notification layer for new filings.
• Kadoa (kadoa.com/layoffs): monthly aggregation, shows Oct 2026 at 2,253 workers across 27 notices (narrower quality filter, do not reconcile against Layoff Atlas without squinting).
• QuiverQuant: ticker-matched, useful for cross-referencing public-company 10-Qs.
• State sources: Mass.gov EOLWD, Colorado CDLE, Washington ESD. Slower, but authoritative.

Pick one for the forward calendar. Use the others only to validate specific filings. Mixing numbers across trackers in a single brief is the fastest way to look sloppy.

WARN is not a news story. It is a 60-day release calendar for a candidate pool nobody else has indexed yet.

Turning the Oct 28 cliff into one actual sourcing brief

The concrete deliverable is a single document per cliff week, listing every filing in your geographic and sector scope, with a shortlist attached to each. For Oct 28 that means roughly 8,220 workers split across employers, filtered down to the ones that match your roles, with a pre-built shortlist already warm.

A workable template:

The payoff of running this exercise once, cleanly, is that the Nov 11 cliff takes half the time: the trackers, the filters, and the outreach cadence are all built. By Q1 2027 you have a repeatable function that treats WARN as inventory, not news.

FAQ

How far ahead can I actually source from a WARN notice?

The WARN Act requires 60 calendar days of advance notice for covered mass layoffs, so from the moment a filing hits the public registry you have up to 60 days before separation and typically another 7 to 14 days before Open to Work banners start appearing. In practice that is a 60 to 75 day window of near-exclusive access, assuming you build your shortlist in the first week after the filing posts.

Should I reach out before someone is formally laid off?

No. Reaching a target while they are still formally employed reads as predatory and burns the lead, and in some cases it exposes them to termination-for-cause risk. Build the shortlist in the pre-WARN window, verify contact details, draft the outreach, and queue it to land 7 to 10 days after the effective date. The whole point of pre-building is that you are ready on day eight, not that you message on day minus thirty.

Which WARN tracker should I actually use?

Pick one tracker as your source of truth for a given window and only cross-reference the others. Layoff Atlas is the cleanest forward calendar for a 90-day view. warnact.io is best for YTD context. QuiverQuant is useful when the employer has a public ticker and you want to validate against 10-Q language. The specific choice matters less than consistency: do not mix a Layoff Atlas 90-day number with a warnact.io YTD number in the same brief.

Does the Oct 28 spike mean layoffs are accelerating?

Not necessarily. The cliff is partly a calendar artifact of WARN's 60-day rule: late-August board decisions around Q3 quarter-end land in late October. One dense week is not a trend. The macro picture is bifurcated, with ADP showing 90,000 private-sector jobs added in September concentrated in health care, education, and hospitality, while white-collar professional services shed jobs. Treat Oct 28 as a sourcing opportunity, not as a market forecast.