Refolk
September 15, 2026·9 min read

128 Companies, 108,459 Layoffs: Your 2026 Target List Just Shrank

The 2026 tech layoff wave is 128 post-IPO names cutting deep, not 1,194 startups cutting shallow. Rebuild your sourcing target list around that.

2026 tech layoffs by companypost-IPO layoffs sourcinglayoffs.fyi September 2026recruiter target list layoffsconcentrated tech layoffs 2026
128 Companies, 108,459 Layoffs: Your 2026 Target List Just Shrank

If your 2026 layoff sourcing playbook still looks like your 2023 one, you are wasting weeks. Second Talent's September 10, 2026 tally of the layoffs.fyi database shows that 84% of this year's tracked tech cuts came from just 128 post-IPO names, and the average company cutting is now shedding roughly twice as many people as it did two years ago. The productive move for headhunters is not more WARN scraping. It is a standing watchlist of about 128 tickers and a plan for the senior engineers walking out of them.

The shape of 2026 layoffs in one paragraph

Fewer companies, deeper cuts, mostly public, mostly US, mostly framed as AI. Through September 10, 2026, layoffs.fyi has logged 128,536 tech employees cut across 299 companies. Of those, 108,459 came from 128 post-IPO firms, meaning 38% of the companies produced 84% of the displaced humans. That is a structurally different market from 2023, when 1,194 companies took smaller bites. Employers named AI in 116,175 US job cuts through August, per Challenger, Gray & Christmas, making it the top stated reason this year.

Here is the comparison your target list should be built on:

Metric20232026 YTD (Sept 10)What it means for sourcing
Companies with tracked layoffs1,19429975% fewer names to watch
Avg employees cut per company~220~430Each announcement is ~2x deeper
Post-IPO share of people cut81% (2025)84% (2026)Concentration is increasing
Post-IPO share of announcementsn/a128 / 333 = 38%38% of names = 84% of people
US share of layoffs.fyi totaln/a103,957 / 128,536 = 81%US-anchored crisis
Repeat cutters (2026 events)n/aAmazon 7, Meta 4, PayPal 46 firms drove 24 events

All layoff-side figures come from Second Talent's parse of the layoffs.fyi public database. Percentages and per-company averages are derived from those totals.

108,459
Tech employees cut from 128 post-IPO companies in 2026 YTD
84% of all tracked cuts came from 38% of the companies. That is your target list.

Why the WARN-filing hustle is now a rounding error

Because the cuts are being made by public-market CFOs to fund AI capex, not by cash-strapped Series B founders, the biggest talent moves are announced through 10-Ks and press releases, not state WARN portals. A single Oracle annual disclosure filed on June 23 moved 21,000 people, roughly 13% of the workforce. Amazon has seven separate layoffs.fyi entries this year. That is more talent per announcement than a diligent WARN scraper pulls in a month of parsing 40-person cuts at private startups.

The mechanism is worth naming, because it tells you where 2027 goes:

  • Public companies are re-underwriting headcount to redirect operating budget into AI infrastructure and AI hires. Oracle's restructuring cost line jumped from $374M to $1.84B in one year, tied to cloud.
  • Private companies with runway are mostly holding pattern or hiring quietly.
  • The people being cut are disproportionately at the senior end, because that is where the compensation dollars are, and compensation dollars are what a CFO trades for GPU contracts.

For a headhunter, the implication is direct. Stop indexing every WARN filing in your state. Start indexing the earnings calendars and annual disclosure filings of the 128 post-IPO names on the list.

The 128-name target list, and the six that matter most

The high-signal move is a persistent watchlist of the six companies that have cut more than three times in 2026, because multi-quarter restructuring programs telegraph the next round. Amazon (7 events on layoffs.fyi), Meta and PayPal (4 each), Salesforce, Uber, and Samsung (3 each) have already told you they are not done. Meta announced 8,000 cuts on May 20 with additional waves projected in H2 2026, on a path toward nearly 20% of the workforce.

The other 122 names divide roughly into three buckets:

  1. One-shot restructurers. Block cut more than 40% of staff in February. Intuit cut 17% in May. Dell shed 11,000 from the hardware segment, taking headcount from 108,000 to ~97,000. These are the highest-volume, lowest-frequency events. Move fast in the first two weeks after announcement.
  2. AI-redirect cutters. Atlassian cut content, support, QA, and project management roles while simultaneously announcing ~800 hires in AI engineering, MLOps, and AI safety. The displaced ICs are candidates for other companies' non-AI backfills, not for their own employer's AI reqs. Vintage matters.
  3. Quiet-cadence cutters. Companies with two announcements this year that you should assume have a third coming inside 90 days.

Sector-wise, retail leads 2026 at 20,200 (mostly Amazon), consumer at 17,544 (led by Meta), and hardware at 14,273 (Dell's 11,000 doing most of it). If you cover any of those three verticals, the target list is even shorter than 128.

The pool you are actually competing for

In Refolk's index of professional profiles, there are roughly 70,822 senior, manager, and director level software engineers currently at US software and internet companies, and that is the pool every sourcer chasing post-IPO cuts is fighting over. That number is the ceiling on what the "shaken-loose" market can produce over the next few quarters, and it explains why open reqs at competitor companies feel harder to fill than the raw layoff numbers would predict.

Two facts sharpen this:

  • The 108,459 figure includes non-engineering roles. Sales, ops, recruiting, and support account for a large share of the total. The engineer-only slice is meaningfully smaller than the headline.
  • Not every displaced engineer is on the market. Severance packages at Meta, Oracle, and Amazon buy several months of "not answering recruiters" behavior. The reply-rate curve peaks 60 to 120 days after the announcement, not the week of.

This is where the sourcing effort has to be surgical, not spray-and-pray. You need to describe the exact person you want at the exact vintage you want them, then have that query run continuously against the index as people update profiles, add "open to work" banners, and move employer fields. That is the exact gap Refolk closes: you describe the person in plain English ("senior backend engineers who left Meta between May and August 2026, Python or Go, based in the Bay Area or Seattle") and get a ranked shortlist back across GitHub, LinkedIn, and the open web.

Geography is misleading, and it changes your outreach

Layoffs.fyi files each cut under the company's headquarters, not where the employees actually worked, which means Oracle's cuts count as American wherever they landed and any India or EMEA target list built off the tracker will undercount by a wide margin. US-headquartered companies account for 81% of this year's tracked layoffs on the site. India is second at 4,725 across 25 layoffs, ahead of Israel at 4,611. The displaced human, though, may sit in Bangalore, Dublin, Tel Aviv, or Seoul.

Three practical consequences:

  • If you source in India, the tracker's own India bucket is a floor, not a ceiling. The real India number is that plus a large slice of the Oracle, Amazon, and Meta totals.
  • If you source in Dublin, the tracker will show almost nothing, because none of the big cutters are headquartered there. LinkedIn's employer-location fields are your source of truth, not the tracker.
  • If you source in Israel, you are looking at a market that punches above its raw ranking because of dense engineering hubs at cut-affected multinationals.

The fix is to run the query against a profile index that carries actual current location, not against the layoff announcement's headline city.

The productive move for 2026 is not more WARN scraping. It is a standing watchlist of 128 tickers.

What to actually do this week

Rebuild your target list from 1,194 fuzzy WARN names down to the 128 post-IPO cutters on layoffs.fyi, then wire persistent queries against each one. Concretely:

  1. Pull the layoffs.fyi CSV and filter to stage = Post-IPO and year = 2026. That is your master list.
  2. Tag the six repeat cutters (Amazon, Meta, PayPal, Salesforce, Uber, Samsung) as "watch weekly." Everyone else is "watch monthly" until an event.
  3. For each name, write out the persona you want in plain English, including function, level, stack, geography, and layoff vintage.
  4. Run those personas continuously, not as one-shot searches. The reply-rate window for a May-cut Meta engineer opens in August, not May.
  5. Match vintage to demand. Displaced non-AI ICs from a company are not candidates for that same company's AI reqs. They are candidates for Company Y's non-AI backfills.

Named entities in the news matter more than aggregate stats when you cold-message. "I saw the Oracle restructuring disclosure on June 23 and figured your team was affected" outperforms "I saw a lot of layoffs are happening" by a wide margin.

The senior-talent inversion

Layoffs in 2026 are no longer a resume-flood problem for junior recruiters. They are a senior pipeline for headhunters, because the cuts have moved up the org chart. Recent reductions include senior positions and specialised technical roles, and the ~430 per-company average is heavy with staff, principal, and manager titles. That inverts the 2023 pattern where post-layoff outreach was mostly ICs with 2 to 5 years of experience.

Stanford's Nick Bloom, quoted on the layoffs.fyi phenomenon, put it plainly: "Having this website engenders more transparency. The stigma has almost totally evaporated." For headhunters running staff-and-above searches, the practical read is that the passive-to-active conversion window at big public cutters is the best it has been in three years. The people are senior enough to know how the game works, tired enough of the current employer to entertain a real conversation, and clustered at a small enough number of firms that a well-built persona query catches most of them on the first pass.

FAQ

How is 2026 different from 2023 for layoff-based sourcing?

The count of companies with tracked layoffs collapsed from 1,194 in 2023 to 299 in 2026 YTD, but the average cut per company roughly doubled from ~220 to ~430. That means the target list is 75% shorter and the per-name yield is 2x deeper. It also means the productive sourcer works a persistent watchlist of ~128 post-IPO names rather than reacting to every WARN filing.

Which companies should be on the standing watchlist?

The six repeat cutters that have already had three or more layoffs.fyi events in 2026 are Amazon (7), Meta (4), PayPal (4), Salesforce (3), Uber (3), and Samsung (3). Beyond those, the highest-signal one-shot events are Oracle (21,000 over 12 months), Block (40%+ of staff in February), Intuit (17% in May), Dell (11,000), and Meta's staged 8,000 in May with H2 waves.

Why is the layoffs.fyi geography data misleading?

The tracker files each event under the company headquarters, not the affected employee's location. US-headquartered firms account for 81% of the 2026 total, but a large share of Oracle, Meta, and Amazon cuts landed in India, EMEA, and APAC offices. To source those markets accurately, run queries against a profile index with current location fields rather than trusting the tracker's country tags.

What is the addressable senior engineer pool right now?

In Refolk's index, there are roughly 70,822 senior, manager, and director level software engineers at US software and internet companies as of September 2026. That is the ceiling on the "shaken loose plus about to be shaken loose" pool that every sourcer running post-layoff plays is competing for. The engineer-only slice of the 108,459 post-IPO cut total is a meaningful but not dominant fraction of that pool.

Try it on the search you came here for

Stop building boolean strings. Just describe the person.

Type one sentence. I plan the search, read GitHub, public LinkedIn and Crunchbase records, and the open web as it is right now, and hand back a ranked list with the reason next to every name.

  1. 01Describe them

    One plain sentence. Role, city, stack, stage, whatever matters to you.

  2. 02I read the web live

    GitHub, public LinkedIn and Crunchbase records, the open web. Not a database that went stale last quarter.

  3. 03You read the shortlist

    Ranked, with the reasoning under every name. Open a profile, ask a follow-up, narrow it down.

  • No boolean, no filters, no seat to buy. One box.
  • Read at search time, so a profile updated yesterday counts today.
  • Every step visible as it runs, every name with its reason.

500 free credits on sign-up. No card, no demo call. See real searches.

Read next