Workday Cut 525 in Product. Its Name Is a Skill on 71,487 US Profiles

Workday's September 29 cut removes about 2.5% of its workforce, mostly in Product and Technology. If you hire HR-tech or fintech engineers, those are the people you want to reach. Then you type "Workday" into a search bar and get back executive assistants, HR staff and the occasional CTO: people who use the product and have never built any of it.

What did Workday announce on September 29, 2026?

Workday is cutting roughly 2.5% of staff. TheStreet puts that at about 525 of 21,070 employees, mostly in Product and Technology, and the company expects $65M to $80M in restructuring charges. Employee actions run until Q1 FY2028, which on Workday's calendar (its fiscal year ends January 31) means roughly February to April 2027.

The numbers, and where they come from:

• Headcount: the roughly 525 figure is TheStreet's own math, based on the 21,070 employees listed in Workday's annual report.
• Charges: $65M to $80M in total, per the 8-K. Of that, $55M to $70M is recognized in Q3 FY2027, and $40M to $55M is cash severance and benefits. Some outlets reported $65M to $85M, but the filing's own components add up to $65M to $80M.
• Timing: the 8-K says actions should be "substantially completed by the first quarter of fiscal 2028, subject to local law and consultation requirements."

The org chart tells you who gets hit. Gerrit Kazmaier (ex-Google Cloud, ex-SAP HANA) became President, Product and Technology in March 2025, and his org took most of this cut. Co-founder Aneel Bhusri is back as CEO for a third time, succeeding Eschenbach.

This is a rotation, not a retreat. Headcount still grew 3% in fiscal 2026 even after the 2025 layoff, the February 2026 round fell mostly on customer operations, and Workday keeps hiring in "key strategic areas." This time the people leaving are the builders.

Why does a "Workday" search return its customers instead of its employees?

"Workday" is both an employer name and a skill, and the people who use the product far outnumber the people who build it. In Refolk's index of professional profiles, 71,487 US profiles list Workday as a skill, and none of the top 25 results worked at Workday.

A keyword collision is when an employer's name is also a product skill, so one search term matches the small group who build the product and the much larger group who use it. Every company running Workday HCM has HR staff and assistants who put it on their profiles. The top skill results were people like that: executive assistants, HR staff and even a CTO, working at Whatnot, Motive, DailyPay and Berkadia.

71,487

Tightening the search helps less than you'd expect. Here is how the US pool breaks down in Refolk's index as of October 2026:

The skill pool is about 88 times the size of the engineer-title pool. Even inside the 808 engineers with "Workday" in their headline, 11 of the 25 sampled worked somewhere else: Wells Fargo, Walmart Global Tech, Microsoft, Meta, Deloitte and KeyBank. One title read "Workday Integration Lead, Architect, and Software Engineer." That is a customer-side engineer who builds on top of Workday. They don't build Workday.

The consultant and analyst group (Workday Consultant, Integration Consultant, HRIS Analyst) is 1.9 times the size of the engineer group. Walmart was the most common employer in that sample with 8 of 25, and consultancies like Kainos, Thoughtworks and Commit Consulting also appeared. Workday didn't make the top 10.

One caveat: those employer shares come from 25-person samples. Read them as direction, not precision.

How do you search by current company instead of skill?

Put Workday in a current-employer field and leave the word out of the keyword box entirely. In a Boolean search, the choice between "current company" and "skill" decides who comes back, because free-text matching can't tell "engineer at Workday" from "Workday analyst at Walmart."

A search that actually returns employees:

1. Filter on current company = Workday. Use a structured employer field, not a keyword.
2. Put engineering titles in the title field. Software engineer, staff engineer, principal engineer, engineering manager.
3. Use internal team names as keywords. Prism, Extend and Adaptive Planning point toward the people who built those products rather than the people who use Workday generally.
4. Exclude customer-side signals. Drop "integration lead," "HRIS," "consultant" and "implementation," and exclude employers that keep turning up as false positives, like Walmart, Wells Fargo and Deloitte.
5. Run a separate past-company search. Past company = Workday, limited to the last 24 months, catches people who left in earlier rounds.

Any free-text search has the same flaw. A query for "Workday" plus "software engineer" brings up every Walmart Global Tech engineer who writes Workday integrations. Without a structured employer filter, the collision wins every time.

That is the exact gap Refolk closes. You describe the person in plain English, including who they are not, and get back a ranked shortlist sorted by where people actually work rather than which words appear on their profile.

How does a 2.5% cut spread out over time?

It plays out over about six months. California's WARN filing covers 142 permanent layoffs in Pleasanton effective November 30, 2026, which is about 27% of the 525. Everyone else leaves on other schedules that stretch to roughly April 2027.

A WARN notice is a legally required advance filing for mass layoffs at a single site above a size threshold. Workday's notice is dated September 29 and covers 6110 Stoneridge Mall Road, the Pleasanton headquarters. As one outlet points out, the public summary does not identify individual job titles or give a complete worldwide location breakdown.

27%

The other roughly 73% fall into three groups:

• Other US states, with their own filings and timelines.
• Sites under the threshold, where too few people are cut to require a notice.
• Offices outside the US. The phrase "subject to local law and consultation requirements" usually points to European works-council processes, and those take months. That is why the 8-K runs all the way to Q1 FY2028.

A realistic timeline:

• Sept 29, 2026: 8-K and California WARN filed.
• Q3 FY2027 (through Oct 31, 2026): $55M to $70M of charges recognized.
• Nov 30, 2026: the 142 Pleasanton layoffs take effect.
• Winter 2026 to 2027: profile updates trickle in from other US sites and smaller offices.
• Feb to Apr 2027 (Q1 FY2028): exits held up by consultation are completed.

The practical point: a single sweep in October misses most of the people. Save the current-company search and run it again every few weeks, because profile updates will arrive over six months, not in one wave.

Is the real talent pool bigger than 525?

Much bigger. Workday has run three reductions in 20 months, so its alumni have been leaving continuously since early 2025, and a past company = Workday search over the last two years finds a far larger pool than this round's headline.

Social posts often quote 8.5% for 2025. The final number in Workday's 10-Q is 7.5%.

The mechanism matters for targeting. Workday keeps hiring in "key strategic areas," and headcount grew overall, so the cuts are a reallocation. Expect them to fall on legacy and non-agent product lines. That means many of these leavers have spent years in the core HCM and financials data model, the hardest part of the stack for a competitor to rebuild.

Pleasanton leavers will be comparing offers against the rest of the Bay Area market.

Why are HCM domain engineers worth more than the 525 headline suggests?

Engineers who know the HCM data model from the inside are rare, and HR-tech rivals are shipping agent products that depend on exactly that knowledge. The headline number counts heads. It doesn't count how scarce their skills are.

Look at what competitors are building. On October 22, Paylocity launched Recruiting Optimizer, which puts "AI agents embedded directly into recruiting workflows" (including Candidate Engagement and Talent Rediscovery agents) and "connects recruiting with onboarding and the employee record." The agent itself is the easy part. The hard part is wiring it into the employee record without corrupting it.

An effective-dated record stores every change to an employee (pay, title, manager, location) along with the date it takes effect, so the system can answer what was true on any given date, past or future. Retroactive pay changes, future-dated promotions and payroll cutoffs all depend on getting this right. An engineer who has shipped on Workday's core objects has already hit the bugs your team hasn't run into yet.

This is also why the consultant pool doesn't substitute. In the index, 1,537 consultants and analysts have "Workday" in their headline, against 808 engineers, but consultants configure the product for customers. They know which screens to change. They don't know how the underlying data model was designed or how it fails.

So filling one of these roles cold is slow. A generic backend search returns thousands of people with no domain depth. A domain search hits the keyword collision and returns customers. The handful who fit are spread across Pleasanton, other US sites and European offices on different exit dates. Founders end up either paying agency fees for a niche search or spending weeks of their own time going through false positives.

How should you pitch laid-off Workday engineers?

Pitch scope and domain fit, not speed. The cash severance pool works out to roughly $76K to $105K per person on average, so the strongest engineers can afford to wait for the right role.

That figure is simple math: $40M to $55M divided by about 525 people. It's an average across levels and countries, so individual packages will vary. The effect is the same either way: "start next week" is not a selling point to someone with months of runway.

What works instead:

• Lead with the domain problem. "We're building agent workflows on top of effective-dated employee records" tells them you know what's hard.
• Name the product area they would own. Engineers from lines Workday deprioritized want ownership, not another reorg.
• Respect consultation timelines. Engineers in countries with consultation requirements may not exit until early 2027. Line up a start date that fits their exit date instead of pushing for an early one.
• Reference their actual work. Mentioning Prism, Extend or Adaptive Planning shows you found them on purpose and didn't just catch them in a keyword sweep.

Severance buys Workday's best engineers time. Win them on the problem, not the start date.

FAQ

How many people did Workday lay off in 2026?

Workday ran two reductions in 2026. The February 2026 plan cut about 2% of the workforce, mostly in customer operations. The September 29 plan cuts about 2.5%, roughly 525 of 21,070 employees by TheStreet's calculation, mostly in Product and Technology. Add the roughly 7.5% cut in February 2025 and that's three rounds in 20 months.

When will laid-off Workday engineers be available to hire?

In waves. The 142 California layoffs in Pleasanton take effect November 30, 2026. Exits at other US sites and in countries that require consultation run until Workday's Q1 FY2028, which is roughly February to April 2027. Expect profile updates to arrive gradually over about six months.

Why doesn't searching "Workday" find Workday engineers?

Because Workday is also a skill. In Refolk's index, 71,487 US profiles list it, and the top results are customer HR staff and assistants, not Workday employees. Even engineers with "Workday" in their headline include many who work at customers like Walmart Global Tech and Wells Fargo. Filter on current or past employer instead of keyword.

Is Workday still hiring after the layoffs?

Yes. Workday keeps hiring in "key strategic areas," and headcount grew 3% in fiscal 2026 despite the 2025 cut. The cuts are a reallocation, which is why the people leaving likely come from legacy or non-agent product lines with deep core HCM experience.