You signed an offer in February. By June your team is gone, your manager is on a PIP, and the CEO sends a memo about "focus." Nobody told you. The signals were all there.
The signals are public. They are filed with state labor departments, posted to SEC EDGAR, logged in commit histories, and printed in trade press 30 to 90 days before HR sends the all-hands. You just have to know where to look — and what counts as a tell.
This is the field guide.
The macro picture: what 2026 actually looks like
Before you panic about every company, get the baseline right.
The U.S. Bureau of Labor Statistics counted 1.9 million layoffs and discharges in March 2026, up 272,000 year-over-year, a layoff rate of 1.2% of nonfarm employment.¹ Through May 2026, LayoffAlert tracked 1,953 WARN Act notices across 42 states, affecting 197,441 employees.² January 2026 alone produced 108,000 announced job cuts — a 118% jump year-over-year and the highest January total since the pandemic, per Challenger, Gray & Christmas.²
That sounds apocalyptic. It isn't, economy-wide.
Fast Company, citing the same Challenger data, notes that overall 2026 announced cuts are running at roughly half of 2025's pace through April — about 300,000 vs. double that at the same point last year.³ Translation: the broader labor market is stabilizing. Tech is not. Roughly 25% of March 2026 announced cuts cited AI as the reason, a sharp jump from prior years.⁴
So the first question to ask about any offer is not "are layoffs happening." Layoffs are always happening. The question is: is this company, in this sector, in this metro, in a hiring posture or a cutting posture right now?
Here is how you answer that in 90 minutes.
Signal 1: WARN Act filings (the legal early-warning system)
The federal Worker Adjustment and Retraining Notification Act requires employers with 100+ employees to give 60 days' written notice before mass layoffs or plant closings. Most states publish these filings in real time.
What the law doesn't do for you: it doesn't cover layoffs under 50 people at a single site, contractor cuts, or "performance-based" terminations. So absence of a WARN filing is not absence of risk.
But presence of one is decisive. If the company you're interviewing at filed a WARN notice in the last 90 days — even for a different office — you treat the offer like a 6-month contract and price your decision accordingly.
Where to look:
- LayoffAlert.org aggregates across 42 states
- Your target company's home-state Department of Labor site (search "[state] WARN notices")
- The federal Department of Labor's WARN database for federal contractors
The fast version is to pull the WARN filings yourself for the company's home state and skim the last 90 days. Funding stage, hiring temperature, layoff signals, recent news — all of it lives in public sources if you know which to read. Thirty seconds beats an hour.
Signal 2: The Challenger language test
Challenger, Gray & Christmas publishes a monthly job-cuts report that categorizes announced layoffs by stated reason: cost-cutting, restructuring, AI/automation, market conditions, M&A, closing.
You use this two ways.
One — benchmark the company's public language. When a CEO says "we are realigning the organization for the next phase of growth," that is restructuring. Restructuring in 2026 means roughly 15% of announced cuts. When they say "we are using AI to operate more efficiently" — that is the 25% AI bucket, and it is the fastest-growing reason. When they say nothing and you find out from Crunchbase News, that is a tell. The language in the press release is the public defense; the WARN filing is the receipt.
Two — sector check. If the company is in a sector where Challenger shows announced cuts running 2-3x the prior-year rate (tech, retail, federal contracting in 2026), your downside risk on the offer is higher than the comp band suggests. The base salary needs to compensate for shorter expected tenure.
Economists at J.P. Morgan and RSM caution that AI-attributed layoffs may be corporate cover for post-pandemic over-hiring corrections rather than genuine automation displacement.³ The honest part: you don't need to resolve that debate. You just need to know that the company is cutting and using AI as the public justification. Either way, your seat is less stable than the offer letter suggests.
Signal 3: Hiring freezes that aren't called hiring freezes
Companies almost never announce hiring freezes. Here is what they do instead:
- Pull job postings without explanation. Use the Wayback Machine (web.archive.org) to compare the company's careers page from 30, 60, and 90 days ago. If 40% of the open reqs vanished and only your role is left, your role is a hole to fill, not growth.
- "Pause" specific functions. Marketing and recruiting go first. If LinkedIn shows the company's recruiting team shrunk by three people in the last quarter, recruiting got cut. The next round is the people they were going to recruit for.
- Slow-roll the loop. Interview-to-offer cycles stretching past 8 weeks for non-executive roles often mean budget signoff is getting kicked upstairs. Crunchbase News documented this pattern across 2024-2026 as a leading indicator of formal cuts.⁵
You check this in 10 minutes: Wayback Machine on the careers page, LinkedIn search for "[company] recruiter" filtered by start date, and Glassdoor interview reviews from the last 90 days.
Signal 4: SEC filings and the funding-stage tell
Public companies disclose. Read them.
- 10-Q "Risk Factors" sections updated mid-year often telegraph cuts coming in the next two quarters.
- 8-K filings for executive departures, especially CFO or COO, in the 6 months before a cycle of layoffs is a documented pattern.
- Cash runway math for private companies: if Crunchbase shows the last raise was 22+ months ago at a Series B-or-later stage, and the company has not announced a new round or revenue milestone, they are inside the burn window where cuts get planned.
For private companies, you triangulate from:
- Crunchbase funding history (date and size of last round)
- Headcount trend on LinkedIn (Insights tab on the company page)
- H-1B disclosures filed with the Department of Labor (signals which functions are being prioritized for retention vs. wind-down)
- GitHub commit velocity on public repos (a 60% drop in commits over two quarters at a tech company is not "focus" — it is fewer engineers)
The full read pulls from all of these — YC API, TechCrunch RSS, SEC EDGAR, GitHub Octokit, WARN Act, H-1B disclosures — and triangulates a hiring-temperature signal. You can do it manually in 90 minutes per company. You should.
Signal 5: The RTO-as-attrition-strategy tell
Return-to-office mandates announced with under 30 days' notice, with no real-estate expansion behind them, are not about culture. They are about attrition without severance.
Crunchbase News documented multiple 2024-2026 cases where RTO announcements preceded formal layoffs by 60-120 days at the same company.⁵ The mechanism: companies cut 8-15% of headcount through voluntary attrition (people who can't or won't commute), then announce the remaining cuts at a smaller, less-publicized number.
If the company you're interviewing at announced an RTO mandate in the last 6 months without simultaneously expanding office leases (check commercial real-estate trade press), you treat the RTO as Round 1 of layoffs.
Signal 6: Performance audits, PIPs, and "calibration"
Three internal signals leak to Glassdoor and Blind within weeks:
- A new performance management system rolled out mid-year
- An unusual number of PIPs issued in Q1 or Q3 (off-cycle from normal review periods)
- "Calibration" meetings that produce more bottom-10% ratings than the prior year
You will not see these in SEC filings. You see them in Blind threads with 200+ comments, in Glassdoor reviews dated within the last 60 days, and in LinkedIn departure waves where 8-12 people from one function leave inside a single month.
A documented pattern, per Crunchbase: companies cut 5-10% through "performance-based" terminations 30-60 days before announcing formal layoffs, because performance-based cuts don't trigger WARN notice requirements at most sizes.⁵
The 90-minute company read (do this before every offer)
Before you sign or counter, run this checklist:
- WARN database — search the company's name and the home state of HQ. Last 90 days.
- Wayback Machine — careers page at 30/60/90 days back. Count the delta in open roles.
- Crunchbase — last funding round date, amount, stage. Compute months of runway at announced burn.
- LinkedIn Insights — headcount trend over the last 12 months. Recruiting-team headcount specifically.
- SEC EDGAR (if public) — latest 10-Q risk factors, recent 8-K filings, executive departures.
- Glassdoor + Blind — reviews dated in last 60 days. Filter for "layoff," "PIP," "RTO," "calibration."
- Trade press — TechCrunch, Crunchbase News, your industry's trade pub. Search company name + last 90 days.
If three or more of those return red flags, the offer is not the offer you think it is. Negotiate accordingly: higher base (because RSUs and bonus are at risk), signing bonus with a short clawback window, severance language written into the offer letter.
How this changes what you ask for
A clean company read changes nothing. A dirty one changes everything.
When the hiring-temperature signals are red, you stop optimizing for upside (equity refresh, promo timeline) and start optimizing for downside protection:
- Base salary up 8-15% to compensate for shorter expected tenure
- Signing bonus with a 6-month clawback maximum — not 12 or 24
- Severance language in the offer letter — "no less than X weeks at the prevailing comp rate" if terminated without cause inside year one
- Equity acceleration clauses — single-trigger on change of control, partial acceleration on termination
This is exactly the kind of read that drives a real counter. 66% of people who negotiate their starting salary succeed — but only 30% even ask.⁶ The 70% who don't ask are leaving money on the table for the same reason they're not reading WARN filings: nobody told them they could.
AMMO reads 1M+ comp data points across 529 role families and 50 metros, refreshed monthly. Grade your offer against the market for your role, level, and metro. The methodology is documented.
What this is not
This is not a way to predict the future. It is a way to read the room.
A company with no red flags can still cut six months in. A company with three red flags can still grow. The point of the read is not certainty — it is to know what you're walking into so you price the offer correctly and write the right protections into the paper.
The corner man does not promise the fight ends in round one. He tells you what the other guy throws.
Run the read
Stop opening offer letters cold. Run the seven-step checklist above first. The thirty minutes you spend is the cheapest leverage you'll ever buy.
Come to the table loaded.
¹ U.S. Bureau of Labor Statistics, "Job Openings and Labor Turnover — March 2026", May 2026. https://www.bls.gov/news.release/jolts.nr0.htm
² LayoffAlert.org, "2026 Layoffs: WARN Act Notices, Statistics & Trends", May 2026. https://layoffalert.org/layoffs-2026
³ Fast Company / Challenger, Gray & Christmas, "Layoffs are actually on the decline in 2026 — but not in tech", May 2026. https://www.fastcompany.com/91538649/layoffs-are-actually-on-the-decline-in-2026-but-not-in-the-tech-industry
⁴ Challenger, Gray & Christmas via ALM Corp, "March 2026 Challenger Report: 60,620 Job Cuts, AI Leads Layoff Reasons", April 2026. https://almcorp.com/blog/march-2026-challenger-report-job-cuts-ai-layoffs/
⁵ Crunchbase News, "Tech Layoffs: US Companies With Job Cuts In 2024, 2025 and 2026", May 2026. https://news.crunchbase.com/startups/tech-layoffs/
⁶ Pew Research Center, "How Today's Workers Feel About Their Job Prospects and the State of the U.S. Economy", April 2023, n=5,775. https://www.pewresearch.org/social-trends/2023/04/13/how-todays-workers-feel-about-their-job-prospects-and-the-state-of-the-u-s-economy/