TL;DR

The first 10 days of May 2026 saw nearly 38,000 US workers cut in tech, finance, media, and travel — including PayPal (4,760 over 2-3 years), Cloudflare (1,100), Coinbase (700), Meta (8,000 effective May 20), and Oracle (30,000+ year-to-date). For recruiters, founders, and investors, this is the largest talent-availability event of the year. ScrapeMaster lets you collect public layoff announcement data, WARN Act filings, and talent signals from layoff trackers — free, browser-based, no API key, no server-side scraping.


The May 2026 wave by company

CompanyCutsFormatEffective date
Meta8,000 (10% of workforce)One-day waveMay 20, 2026
Cloudflare1,100 (20% of workforce)One-day waveEarly May 2026
Coinbase700 (14% of staff)One-day waveEarly May 2026
PayPal4,760 (20% of staff)Multi-tranche, 2-3 yearsRolling
Oracle30,000+ YTDMulti-trancheRolling
Fidelity800 (with simultaneous hiring)RestructuringQ2 2026

YTD total tech industry: 113,863 workers across 179 layoff events (per Skillsyncer). AI cited as the reason for 49,135 cuts, accounting for ~16% of all job-cut plans in 2026.

For anyone whose work involves the tech talent market, this is a once-a-year intelligence opportunity. The amount of public information is huge — WARN Act filings, press releases, layoff trackers, company blog posts, executive interviews — and it's distributed across dozens of sources.

A structured scraping workflow turns that distributed signal into a usable dataset.


Who needs layoff intelligence

Recruiters & talent acquisition

Layoff announcements identify talent pools before those candidates start applying broadly. The information available days before the candidates flood LinkedIn:

  • Which divisions were cut (engineering, product, sales, ops)
  • Seniority levels (often disclosed in press coverage)
  • Geographic concentration (WARN filings name specific facilities)
  • Effective dates (when candidates are technically available)
  • Severance package terms (relevant to candidate timing decisions)

A recruiter who pulls Meta's Burlingame WARN filing on Monday has 124 candidates' job titles and effective separation dates available — three days before those candidates have updated their LinkedIn.

Startup founders

Large-company layoffs are talent acquisition opportunities. A laid-off senior engineer from Meta or Coinbase brings:

  • Institutional knowledge (engineering systems at scale)
  • Network (existing connections to other strong engineers)
  • Available capacity (immediate, not waiting on notice periods)
  • Compensation expectations rebased (especially if equity took a hit)

Beyond hiring, the pattern of layoffs reveals strategic shifts. Meta cutting Reality Labs in January, Facebook in May. PayPal trimming over 2-3 years. Oracle cutting deep. Each is a signal about where competitors are pulling back, which can be a market opening for a startup focused there.

Investors & analysts

Layoff announcements are material corporate events:

  • Free cash flow implications (severance cost vs. ongoing salary savings)
  • Strategic direction (which divisions get cut tells you which got prioritized)
  • Sector-level patterns (AI restructuring vs. cost-cutting vs. revenue softness)
  • Competitive positioning (who is hiring while peers are cutting)

A scraped, structured dataset of 100+ layoff announcements with division-level breakdowns is the kind of thing equity research analysts pay $$$$ a year for. It's also assembleable from public sources with a Chrome extension and a few hours.

Job seekers

If you're at elevated layoff risk yourself (software development, mid-level tech management, fintech), tracking patterns helps you:

  • Calibrate your own risk
  • Identify companies that are hiring while peers cut
  • Time your search to land before the next 8,000-person wave saturates the market
  • Negotiate from a position of better information

What's scrapeable, and where

The public surface area for layoff intelligence is rich:

SourceWhat you getUpdate cadence
Skillsyncer Layoffs TrackerYTD events, company-by-companyReal-time
TrueUp LayoffsAggregated event logReal-time
informationweek.com 2026 listEditorial summary by companyWeekly
24/7 Wall St / Yahoo FinanceSector-level analysisPer news cycle
State WARN Act portals (CA EDD, NY DOL, TX TWC, etc.)Required filings by companyPer-filing
Company press releases (PayPal IR, Meta newsroom, Oracle IR)Official announcementsPer-event
LinkedIn (logged in, your own session)Specific candidate signalsReal-time
Company blogs (Cloudflare blog, etc.)Executive narrativePer-event

Tools that work well for this stack:

  • ScrapeMaster — Chrome extension, click on a layoff tracker page and export the table to CSV
  • Manual review — for press releases, executive interviews
  • State WARN portals — typically public web tables, scrapeable with the same workflow

You don't need a paid API service. You need a structured approach to the public web.


A practical workflow for the May wave

Day 1 — Build your watchlist

Decide which companies you care about. For most use cases:

  • The four headline names (PayPal, Cloudflare, Coinbase, Meta)
  • Any company in your niche (fintech, e-comm, infra, AI, dev tools)
  • The big aggregators (Skillsyncer, TrueUp, informationweek)

Open each in a tab, install ScrapeMaster.

Day 2 — Scrape the aggregators

Visit Skillsyncer Layoffs Tracker. The page has a structured table of events by company. Click ScrapeMaster, select the table region, export to CSV. Repeat for TrueUp and informationweek's list.

You now have three overlapping datasets with company, date, headcount, and (sometimes) division. Cross-reference for accuracy.

Day 3 — Scrape state WARN filings

California EDD, New York DOL, Texas TWC, and similar state portals publish WARN Act filings — the legally required 60-day notices for mass layoffs. Each has a public web view. Scrape the table for filings in the last 30 days.

You'll find:

  • Filer (company)
  • Facility (street address)
  • Effective date
  • Number of positions
  • Job titles (sometimes, depending on state)

This is the most reliable source for upcoming layoffs — companies must file 60 days in advance, so the WARN data leads the news cycle.

Day 4 — Scrape press releases and executive interviews

Each company posts its own version. Meta's newsroom, PayPal's investor relations page, Cloudflare's blog. The narrative ("we're investing in AI," "we're streamlining operations") is useful context but isn't the structured data you need.

Day 5 — Cross-reference with LinkedIn

This step is manual and uses your own LinkedIn session. Don't run automated scrapers against LinkedIn. Instead, browse logged-in as yourself, search for specific titles at the affected companies, and use ScrapeMaster to structure what's on your screen.

The legitimate use case: a recruiter reaching out to specific candidates. Not a server-side bot harvesting 500,000 profiles.


A sample dataset structure

After a week of work, your CSV looks something like:

CompanyDateHeadcountDivisionSourceEffectiveSeveranceState
Meta2026-04-238,000Cross-orgPress release2026-05-2016w + 2/yCA, multi
Meta-Burlingame2026-04-22124UnknownWARN2026-05-22per packageCA
Cloudflare2026-051,100Cross-orgPress release2026-05per packageCA, multi
Coinbase2026-05700Cross-orgPress release2026-05per packageCA
PayPal2026-Q24,760RollingIR releaseRollingTBD per trancheMulti-state
Oracle2026 YTD30,000+Multi-tranchePress, WARNRollingper packageMulti-state
Fidelity2026800Tech restructuringPressQ2per packageMA, TX

This is a dataset you can pivot, filter, and act on. Recruiters can sort by state for outreach. Founders can filter to specific divisions. Investors can compute headcount-weighted sector totals.


ScrapeMaster vs the alternatives

ToolArchitectureSetupCostLinkedIn ToS risk
ScrapeMasterChrome extension, in-sessionOne clickFreeLow (manual session)
OctoparseDesktop / cloud workflowsLearning curveLimited freeMedium-high (cloud)
ParseHubDesktop / cloudLearning curveLimited freeMedium-high (cloud)
ApifyCloud platform with scrapersCode/configurationPay per useVaries by scraper
Import.ioEnterprise SaaSAccount + setupPaidVaries
SimplescraperChrome extensionClick-basedLimited freeLow
Web Scraper.ioChrome extensionClick-based, complexLimited freeLow
Custom Python (requests + bs4)DIYHours-daysFree (your time)High if run server-side

For "scrape three layoff trackers and a state WARN portal in one afternoon," ScrapeMaster is the lowest-friction choice. The cloud-based options give you more power for unattended runs but invite legal exposure if you point them at LinkedIn or aggressively at any site that doesn't want bots.


What to do with the dataset

Once you have a clean dataset, the actions diverge by role:

Recruiters: build a per-state outreach plan; cross-reference job titles to your open roles; prioritize candidates from divisions cited as cut in the press release.

Founders: build a per-division target list; identify specific engineers or PMs from the cut divisions who'd fit your team; reach out with concrete role descriptions, not generic "interested in joining?" pings.

Investors: aggregate to sector level; compute headcount-weighted YTD layoffs by sector; correlate with stock-price movements and equity research notes.

Job seekers: identify companies hiring in your niche while peers cut; prioritize applying to those before the market saturates; benchmark severance expectations against publicly disclosed terms.

For all of these roles, archiving the source pages with Convert: Web to PDF on the day you scrape is a useful belt-and-suspenders move — sources change, archives last.


Compliance and ethics

A few rules that apply regardless of role:

  • Don't scrape LinkedIn aggressively — your own session is fine, automated server-side scraping is not. The recent litigation wave (HiQ remand, ongoing 2026 cases) and the Reddit v. Perplexity / SerpApi / Oxylabs / AWMProxy lawsuit make this clear.
  • WARN Act filings are public — they're filed precisely because the law requires public disclosure. Treat them as such.
  • Press releases are public — collect freely.
  • Don't reach out by name to laid-off individuals without their consent — let them update LinkedIn first.
  • Don't share the dataset publicly if it contains identifiable individuals before they've publicly announced status.

A note on AI tools for layoff analysis

You'll see plenty of "AI talent intelligence" SaaS pitches in May 2026. Most of them are buying data from the same scraping intermediaries currently being sued. For sensitive work, build your own dataset with ScrapeMaster, and run any AI analysis locally or on infrastructure you control.

If you need to compare which AI models are best for the analysis layer (Claude, GPT, Gemini), CineMan AI gives a side-by-side view without uploading any of your data.


Frequently asked questions

Q: Are WARN Act filings really public?

Yes. The federal WARN Act requires 60 days' advance notice of mass layoffs (generally 50+ employees at a single site) and individual states maintain public registries of those filings. California's EDD, New York DOL, Texas TWC, and most other states publish them online.

Q: How quickly do laid-off employees update LinkedIn?

The pattern across 2025-2026 waves: 30–50% within 72 hours, 70–80% within two weeks. The "early days" window is where WARN data and press releases give you a lead over LinkedIn.

Q: Can I scrape Skillsyncer's tracker directly?

Their tracker page is a standard web page with structured content. ScrapeMaster can extract the visible table. As always, respect their terms of service and don't hammer their servers — for periodic scraping, once per week is plenty.

Q: What about layoffs.fyi?

layoffs.fyi has historically been the de facto aggregator. Similar approach applies — visit, select the table, export. Cross-reference with Skillsyncer and TrueUp for completeness.

Q: Is there a risk of getting blocked by aggregators?

If you hit a single page once per day, no. If you fire 1,000 requests at the same domain from a server, yes. ScrapeMaster runs in your browser at human speed, which is the right rate.

Q: How do I structure the CSV for CRM ingest?

Most CRMs ingest CSV with standard columns: Name, Title, Company, Location, Source, Notes. For layoff intelligence, add a "Layoff Event Date" and "Available From" column. Pipedrive, Salesforce, HubSpot, and Close all accept this format.

Q: What about international layoffs?

WARN-equivalents exist in many jurisdictions (UK collective consultation rules, EU social plans). Scraping public filings in those jurisdictions is the same workflow, different source. Your country may also have public business registries that disclose mass layoffs.

Q: Does Meta's 16-weeks-plus-tenure benchmark apply elsewhere?

Meta's package is one data point — not a market floor. Cloudflare, Coinbase, PayPal, and Oracle haven't all published their packages publicly. Where you can collect them (often from anonymous Blind posts, with appropriate skepticism), tag the source explicitly.

Q: Should I save the source pages as PDF too?

Yes. Use Convert: Web to PDF on each scraped page the day you scrape. Sources update; archives stay frozen.

Q: How long is the wave likely to last?

The pattern across 2024-2026 suggests waves persist 6-12 months once they start. AI-driven restructuring is a multi-year story, not a one-quarter story. The companies you watch this May will still be reshaping their workforce in 2027.

Q: Can I share the dataset with my team?

Yes — internally. Don't post identifiable individual-level data publicly. Aggregate counts and division-level cuts are fair game; "John Doe was laid off from Meta on May 20" is not, until John publicly says so himself.

Q: What if I want to scrape salary data?

Levels.fyi, h1bdata.info, and state-specific salary disclosure portals are public sources. Same workflow applies. For California specifically, the California Pay Data Reports portal publishes aggregated data by company.


Bottom line

The May 2026 layoff wave is the biggest talent-availability event of the year. The public data needed to make sense of it — WARN filings, press releases, tracker sites — is huge and free.

ScrapeMaster turns that distributed signal into a structured dataset in an afternoon, with no API keys, no server-side scrapers, no LinkedIn enforcement risk. Pair it with Convert: Web to PDF to archive sources, and you have a defensible intelligence workflow that holds up regardless of where the 2026 wave goes next.