Workday Just Cut 525 P&T Engineers. The 8-K Is Your Target List.

On September 29, 2026, Workday filed an 8-K disclosing a roughly 2.5% workforce cut, "primarily within Workday's Product and Technology team." No WARN notice has posted yet, so the names are not public. But the filing itself, read correctly, is tighter than any WARN list a California sourcer will see this quarter.

What the 8-K actually says, and why it beats a WARN notice

The 8-K pins the cut to a specific segment, a specific HQ, and a specific completion window: ~525 Product and Technology roles, mostly Pleasanton, substantially done by Q1 FY28. That is enough to start sourcing on day one, which is the whole point.

Here is the filing in numbers:

• ~2.5% of a ~21,000-person workforce, which works out to approximately 525 roles.
• $65M to $80M in total charges, of which $40M to $55M is cash severance and benefits.
• $55M to $70M booked in Q3 FY27, ~$10M in Q4 FY27, with ~$15M in non-cash lease impairment on leased office space Workday is giving back.
• Employee actions "substantially completed" by Q1 FY28, subject to local law and consultation.
• Primary segment: Product and Technology. Not GTM, not Customer Operations, not Finance.

Compare that to the February 2026 round, which cut roughly 400 roles in Global Customer Operations. That one did produce a California WARN, naming 154 Pleasanton employees with a last day of April 6, 2026, roughly 61 days after the filing. If September's cut follows the same cadence, the last day of work lands somewhere in late November or early December 2026.

The trap is waiting for the WARN. California's WARN Act will eventually produce a list, but only for California workers, and only after a 60-day clock has started burning. The February WARN captured 154 of ~400 cuts, about 38% of the population. The other 62% were sourced by competitors who read the 8-K and moved. The 8-K gives you the segment on day one, and segment is the field you actually filter on.

The index numbers nobody else will publish

In Refolk's index of professional profiles, only ~163 US senior, staff, principal, or EM-level software engineers list "Workday" as a skill on their public profile. That is a vanishingly thin pool against a US senior-engineer base of roughly 704,500.

The implication is uncomfortable for anyone treating this as a volume play. The 525 engineers being cut are more than three times the entire publicly-identifiable Workday-skilled senior-engineer pool. They are not interchangeable with "any backend engineer in the Bay Area." They are among the only people on earth who have shipped code inside the Workday platform itself.

322%

The second number worth staring at is the composition of that 163. The current-employer distribution is dominated by Workday customers, not Workday alumni: Wells Fargo (8), Walmart (2), KeyBank, and other large enterprises running Workday internally. Charlotte, NC and Bentonville, AR show up as geographic clusters for the same reason. If you were planning to source "ex-Workday engineers" by filtering LinkedIn skill tags, you were going to call Wells Fargo IT. The actual builders, the 525, barely surface in that search at all until their profiles update post-exit.

Why Rippling and Deel will outbid Wells Fargo

Workday-skilled talent is almost all buy-side. The 525 being cut are the sell-side, and sell-side HCM engineers price completely differently from the IT staff who configure Workday at a bank.

Rippling, Deel, SAP SuccessFactors, Oracle HCM Cloud, and Gusto all need people who have shipped inside a multi-tenant HCM platform, not people who have configured one. That is a tiny labor market. A senior engineer who built a Workday integration framework, or EIB, or Studio tooling, or the core HCM object model is worth materially more than a Wells Fargo Workday admin commands, because the skill is architectural and the supply is 525 wide for the next 60 days.

The named buyers you should assume are already in the market:

• Rippling. Direct HCM competitor, SF-based, well-funded, known for poaching from incumbents. Will pay in cash and RSUs, in-person expected.
• Deel. Global payroll and EOR, remote-first by policy. The remote offer is itself worth 10 to 15% in effective comp against a Pleasanton RTO.
• SAP SuccessFactors. Needs exactly the "legacy" platform engineers Workday is cutting, not the AI generalists Workday is keeping.
• Oracle HCM Cloud. Same thesis as SAP. Core HCM data model experience is the exact ask.
• Gusto. SMB payroll moving upmarket, SF HQ, geographically native to Pleasanton commuters.

The non-obvious counter-bidders are the Workday customers themselves. Wells Fargo in Charlotte and Walmart Global Tech in Bentonville already run the biggest concentrations of Workday-skilled senior engineers in the Refolk index. They will try to hire some of the 525 as internal platform leads. They will lose most of them on comp and on the "I did not spend 7 years building Workday to go configure it at a bank" argument, but they will anchor the floor.

The AI reorg tell: who is actually in the 525

The cut skews toward legacy platform engineers, not AI or ML. Workday said plainly it will "continue to hire in key strategic areas" throughout fiscal 2027, and the company just rebranded as "the enterprise AI platform for managing people, money, and agents." Aneel Bhusri came back as CEO on February 9, 2026 explicitly to run that pivot, replacing Carl Eschenbach.

Read that way, the segment inside the segment matters more than the segment itself. If you are sourcing the 525, bias toward:

• Core HCM object model engineers (worker, position, job, compensation).
• Integrations: Studio, EIB, Workday Web Services, Core Connectors.
• Reporting and Prism Analytics backend.
• Platform infrastructure predating the current Illuminate and agents push.
• Managers and tech leads whose titles have not changed in 3+ years. The reorg hits them first.

Bias away from, because Workday is retaining them:

• Illuminate, agents, and the AI/ML platform.
• Anyone who moved to a "GenAI" team in the last 18 months.
• The recently-promoted. Fresh promotions are a retention signal.

The practical implication for Oracle HCM and SAP SuccessFactors is that this is almost gift-wrapped. The exact skill set they need, core HCM platform engineering, is the exact skill set Workday just told the SEC it is cutting. For Rippling and Deel, the AI engineers are harder to pry loose, but the integrations and extensibility people are gettable, and those are the ones who make a competing HCM platform usable by enterprises.

The 525 being cut are more than three times the entire publicly-identifiable Workday-skilled senior-engineer pool.

The 90-day window, not 60

Your real hiring window is closer to 90 days, not the 60 that WARN implies. Severance math is why.

At $40M to $55M in cash severance across 525 heads, the midpoint is ~$90K per engineer. Add stock-based comp acceleration and benefits continuation and the typical P&T engineer has 3 to 5 months of runway without touching savings. That means:

1. Offers that land in October 2026 convert at the highest rate. The engineer has heard the news, has not yet decompressed, and wants certainty before severance starts.
2. Offers that land in November to December 2026 compete with holiday inertia and family logistics but still convert well, especially remote-first offers.
3. Offers that land in February 2027 compete with re-vested RSUs at the new employer, a working spouse's opinion, and a 90-day gap on the resume the candidate has started to rationalize. Conversion falls off a cliff here.

The practical sourcing cadence is: identify in October, make first contact in October, interview through November, close before Thanksgiving. If your team is still "building the list" in December, you are hiring the second-choice 525, not the first.

Pleasanton geography is a chokepoint, not a search filter

Pleasanton is where almost all of these engineers physically are, and with Bhusri back and in-office pressure up, Workday has been pulling P&T toward the HQ campus, not away from it. That matters in two directions:

• For in-person buyers (Rippling in SF, Gusto in SF, Oracle in nearby Redwood Shores), Pleasanton is a 45-minute drive and candidates will take meetings.
• For remote-first buyers (Deel, and remote teams at SAP), location flexibility is itself a differentiator worth roughly 10 to 15% in effective comp against a continued Pleasanton RTO.

The segment is small enough that you should assume every serious competitor is working the same list. Differentiation on offer structure, not sourcing speed, is where this gets won in week six.

How to actually build the list in week one

Build it from the 8-K segment downward, not from the eventual WARN upward. The segment, the HQ, and the sub-team are the three axes that produce a usable list on day one.

A workable week-one process:

1. Pull everyone with current title at Workday, Pleasanton (or Bay Area), Product and Technology org, hired on or before early 2024. Tenure cutoff removes the Illuminate pivot hires who are being retained.
2. Layer skill signals: Scala, Java, Workday Studio, EIB, Workday Web Services, Core Connectors, Prism, PICOF, workday-studio GitHub repos. Avoid filtering on "Workday" as a LinkedIn skill, because the Refolk index shows that tag mostly surfaces customer-side admins at Wells Fargo and Walmart.
3. Cross-reference GitHub activity. The 525 includes engineers who contribute to internal tooling that occasionally mirrors to public repos. Low commit activity on public GitHub but a long tenure at Workday is often the right signal, not the wrong one.
4. Rank by manager tier. ICs convert faster than managers; staff and principal ICs are the highest-leverage hires because they bring architectural judgment Rippling and Deel cannot easily buy.
5. Save the search. Re-run weekly. Profile updates ("Open to work," title changes, location edits) will spike in the two weeks after the last day of work in late November.

Describe the segment in a sentence ("P&T engineers at Workday in Pleasanton, 3+ years tenure, core HCM or integrations, not on AI teams") and Refolk returns a ranked list across GitHub, LinkedIn, and the open web, without you writing a Boolean. When the WARN finally posts in December with its 150-ish California names, you will already have offers out.

FAQ

How many Workday employees were laid off in September 2026?

Approximately 525, or about 2.5% of a ~21,000-person global workforce, per the 8-K filed September 29, 2026. The reduction is "primarily within Workday's Product and Technology team." Combined with the February 2026 cut of ~400 roles in Global Customer Operations, Workday has reduced headcount by roughly 925 people, or ~4.4%, across the year.

Will there be a California WARN notice for the September 2026 Workday layoffs?

Almost certainly yes, because California's WARN Act's only statutory exception is a physical calamity or an act of war, which does not apply here. The February 2026 round produced a Pleasanton WARN with 154 names and a last day of April 6, 2026, roughly 61 days after disclosure. Applied to the September 29 announcement, expect a WARN posting with a last day in late November or early December 2026. The WARN will likely cover 30 to 40% of the total cut, mirroring the February ratio.

Who is best positioned to hire ex-Workday Product and Technology engineers?

Rippling, Deel, SAP SuccessFactors, Oracle HCM Cloud, and Gusto are the primary buyers, and they are not interchangeable. Rippling and Gusto want in-person Bay Area engineers and will match on cash. Deel wins on remote flexibility, worth 10 to 15% in effective comp against a Pleasanton RTO. SAP SuccessFactors and Oracle HCM Cloud want the legacy core-HCM and integrations engineers specifically, which is exactly the cohort Workday is cutting rather than retaining.

Why is sourcing "Workday skill" on LinkedIn the wrong approach?

Because the public pool of US senior engineers listing "Workday" as a skill is only ~163 people in the Refolk index, and it is dominated by customers like Wells Fargo (8) and Walmart (2), not by Workday builders. The 525 being cut are platform engineers whose public profiles use language like "Scala," "distributed systems," "integrations platform," or "HCM" without ever naming the product. Filter on segment, tenure, and location from the 8-K, not on the skill tag.