The Layoff Watchlist Playbook: Reach Affected Talent First

This is the operating manual for a recruiter or sourcer who wants to reach affected people the moment a target company starts shedding the exact roles they hire. It is for in-house recruiters, sourcers, talent leaders, and founders doing their own hiring. It turns scattered public signals - WARN filings, review-rating drift, hiring-freeze and exec-departure news, open-to-work clusters - into one standing process you run every week, so you are ahead of every recruiter who waits for the news.

Most public writing on this topic is either advice on being empathetic to laid-off candidates or worker-facing "am I about to be fired" posts. Neither gives you a repeatable operation. This one does: a maintained watchlist, signals ordered by lead time, a weekly ranking that names the company to work now, and a lawful send.

Why a standing watchlist beats reacting to the news

A watchlist wins because the useful signals arrive in a predictable order, and the earliest ones arrive weeks or months before the layoff list does. Companies rarely announce layoffs without warning. Financial pressure signals appear first, then operational changes, then HR activity. If you only react when a name hits a layoff tracker, you are competing at the last stage of that sequence with everyone else who reads the same tracker.

The math also forces the point. Small layoffs under 50 people rose from 38 percent of layoff events in 2015 to 51 percent in 2025. Federal WARN only triggers on a mass layoff of 50 or more at a single site, so more than half of all events by count never touch the statutory feed at all. A watchlist that reads only WARN notices looks quiet while real cuts happen inside your target companies.

There is a counterweight worth knowing. Where the cuts are large, they concentrate: 2025 cuts spanned fewer companies (257 versus 551 in the prior comparison), signaling concentration among larger employers. Fewer, bigger events mean a tightly scoped target list can still cover most of the affected talent volume. So the watchlist is not a firehose. It is 30 to 80 named companies, watched in order.

The earliest-mover advantage lives in leadership changes and budget language, not in layoff lists everyone reads at once.

The signals, ordered by lead time

Rank your signals by how far ahead of the announcement they fire, because that ordering is the whole edge. Exec departures and cost freezes give you weeks to months. WARN filings and open-to-work banners give you days or none. You work the early signals to build the shortlist and the late signals to time the send.

The exact week-by-week lead time per signal is not established publicly beyond the ranges below, so treat these as ordering, not a schedule. A new CEO, CFO, or division head is one of the earliest and most reliable signs, with new leadership restructuring within 60 to 90 days in most cases. Cost-control measures often precede layoffs by 30 to 90 days, because leadership exhausts less painful cost-reduction methods first, then turns to headcount when those fall short.

Table C - Signal lead time ordering (published ranges)

Practitioner accounts fill in the middle of the sequence: hiring freeze first, then leadership departures and cross-org reorgs, then rescinded offers, PIP activity, and rumors, with travel bans as a late-stage sign. Every signal proves something different, and each one lies in a specific way. Know both before you act on it.

Glassdoor is the trickiest of the four. Ratings drop immediately after a layoff and take more than two years to recover, so the headline rating registers at or after the event, not before. As an early predictor it is a false positive. Use leadership-subrating drift as a weaker leading hint only, and lean on it hardest at prestige employers, where the signal is loudest.

The public data sources and how fresh each one is

Your hard feed is the state WARN portal; your soft feeds are review-rating drift, job-post takedowns, exec news, and open-to-work clusters. The critical fact about WARN is that no single national database is authoritative. The U.S. Department of Labor neither maintains a database of WARN notices nor requires employers to file them with the Department; employers file with the state dislocated worker unit. State publication is voluntary and varies, so the frequency of listings and the amount of information differ from state to state.

That means you assemble the feed yourself, one portal at a time, plus one aggregator to catch what individual portals miss. Aggregators help: one searchable 50-state WARN database holds 82,000-plus filings covering 8.87 million-plus workers. Live tech trackers refresh continuously; two such pages were both refreshed within 20 hours at time of research. Where a state lags, file a public-records request. California, for example, processes public-records WARN requests within 10 days.

One timing note that trips people up: filings may be made available before or after a public announcement, so following both the statutory portal and the live trackers is advised. Neither source alone tells you the true moment. Watch both.

Run the watchlist: the weekly procedure

Here is the operation start to finish. The one-time setup is steps one through four and takes a few hours; steps five through eight are the weekly and monthly rhythm. Roles and rough durations are noted so a talent leader can staff it.

The scoring in step five is deliberately simple: a company rises when it has more active signals, fresher signals, and earlier signals. A company with a fresh exec departure and a hiring freeze outranks one with only a two-month-old Glassdoor dip. The output is a single ranked "work-this-week" line-up, not a dashboard you admire.

The hardest part of the whole operation is step seven: converting a fired signal into named, contactable people who match your open role, fast, before the open-to-work surge draws every other recruiter. That is where a plain-English search across public profiles collapses hours of manual cross-referencing.

Refolk is where I run that intersection query without stitching a WARN list to a profile export by hand. When a filing fires, Refolk turns the affected company into a ranked list of people in the roles you fill, which is the difference between working the signal this week and finding the same people after they have thirty open conversations.

Where this goes wrong: failure modes and false positives

This section is the most valuable part of the playbook, because most of the ways a layoff watchlist fails are quiet. It keeps running, it produces names, and the names are wrong or the sends are unlawful. Here are the failure modes from the field and the check for each.

• WARN-only coverage gap. Over half of 2025 events were sub-50 and never trigger federal WARN, so a WARN-only list looks empty while cuts happen. Check: always pair WARN with Glassdoor drift and open-to-work clusters.
• Reorg mistaken for layoff. Some companies keep hiring while cutting positions; restructuring is a strategic realignment, not always a cut. False positive: an org-chart change with no headcount drop. Check: confirm a net headcount decline before you work the company.
• Glassdoor lag. Ratings drop immediately after a layoff and take more than two years to recover, so they register at or after the event. Using them as an early predictor is a false positive. Check: use leadership-subrating drift only, as a weak leading hint.
• Open-to-work is late. The banner fires at or after separation and gives zero lead time. Check: treat it as a "send now" trigger, never a "predict" signal.
• Recruiters-only invisibility. Recruiter-only open-to-work status is hidden from non-recruiters and visible only inside LinkedIn Recruiter tools, so sourcing off the public green banner alone misses discreet job-seekers. Check: do not assume the visible banner is the whole affected pool.
• Unlawful mass outreach. Blasting mass messages without a specific role in mind is likely unlawful, and candidates must be able to opt out easily. Check: one genuine req per outreach batch, documented.
• Stale or hoarded data. Building a talent database "in case you need it later" is not lawful under GDPR. Check: enforce a contact-or-delete rule so records do not sit past their basis.

The prestige effect sharpens where the Glassdoor signal is real. Top-tercile-rated employers drop 0.22 stars in the first six months post-layoff, versus 0.02 for bottom-tercile employers - roughly an 11x gap. So sentiment is a strong signal at well-branded companies and near-useless at already-low-rated ones. On average, layoffs cut ratings by 0.13 stars, with surviving employees dropping 0.16. Use the subrating drift, and weight it by the employer's baseline.

The compliance layer that keeps the watchlist running

Treat lawful basis as the moat, not the friction: a documented legitimate-interest basis per role is what lets you run an always-on watchlist while unlawful scrapers get fined. WARN notices themselves are public records - Colorado treats them as public records under its open-records act - so reading the filing is not the risk. What you do with the person afterward is.

In the EU and UK, sourcing runs on legitimate interest with duties attached. UK ICO guidance says legitimate interest is likely to apply if you are recruiting for a genuine vacancy with minimal intrusion, but blasting mass messages without a specific role in mind is likely unlawful. Transparency is mandatory: send sourced candidates an email informing them you are processing their data within one month of first processing it, link to your privacy notice, and if you miss that window, delete their data.

Enforcement is not theoretical. The Irish Data Protection Commission fined LinkedIn EUR 310 million in October 2024 for relying on invalid consent and unlawful legitimate-interest claims for behavioral profiling. The lesson for a watchlist is direct: tie every outreach batch to a real, open requisition, and document it.

Requisition: <live req ID and role title>
Signal that triggered this batch: <WARN filing / open-to-work surge / exec departure>, dated <date>
Source of candidate data: <state WARN portal / aggregator / public profile>
Basis: Legitimate interest - recruiting for a genuine, open vacancy; minimal intrusion.
Transparency step: Processing notice + privacy-notice link sent within 30 days of first processing.
Retention: Contact-or-delete. Delete if no engagement and no active req within 30 days.

Where your leverage is, by market and skill

Refolk's index makes the timing argument concrete, and it explains why "reach them first" matters more in some markets than others. In Refolk's index there are 21,309 profiles in the US with recruiter or sourcer titles versus 729 in the UK - about 29 times more. In a thin market, being first to an affected candidate matters more, because there are fewer competing sources to absorb your slippage. The same late send that costs you nothing in a deep pool costs you the candidate in a shallow one.

Table A - Recruiter/sourcer talent pool, US vs UK (Refolk's index)

Supply also varies sharply by the skill you hire for, which changes how much a single layoff event moves your pipeline. In Refolk's index, 13,775 US Software Engineers list Kubernetes versus 5,401 listing React - about 2.55 times as many. When you recruit a scarcer skill, each affected company is worth more to you, and the case for early, precise timing gets stronger.

Table B - Skill supply among US Software Engineers (Refolk's index)

Use these numbers to set your own posture. Thin market or scarce skill: prioritize speed and lean hard on the early exec and freeze signals so you are drafting before the open-to-work surge. Deep market or common skill: you can afford to wait for confirmation signals and work volume.

Keeping the watchlist current

A watchlist decays the moment you stop tending it, so build the monthly prune into the rhythm rather than treating it as cleanup. Targets get acquired, roles you fill change, and false positives accumulate. Run the checklist below before you call any weekly cycle done, and run the full prune monthly.

To keep the sources themselves fresh, re-check each state's WARN publication cadence quarterly, since publication is voluntary and a state can change what it posts. Re-verify your aggregator still refreshes on the schedule you assumed. And revisit your target list every quarter against the roles you are actually filling, because a watchlist aimed at last quarter's hiring plan quietly stops earning its keep. The operation is only as good as the last time you tended it.