Upwork Is Cutting 37%. Its 1,600 Freelancers Aren't in the Count
Upwork's 37% cut is roughly 230 staff on its last base, plus an unnamed subsidiary. Its 1,600 internal freelancers aren't counted. How to find each group.
Upwork's amended 8-K raises its planned workforce cut from about 24% to about 37%, but it gives a percentage, not a head count, so anyone sourcing from it is working from an estimate. On the last headcount Upwork disclosed, 37% comes to roughly 230 people, and that figure leaves out the roughly 1,600 independent workers Upwork used on its own internal projects in 2025. If you hire engineers, three separate groups are coming onto the market, and you'll find each one in a different place.
What did Upwork actually announce in the amended filing?
Upwork filed an amended 8-K expanding its restructuring to about 37% of total workforce. That figure includes the 24% previously disclosed plus a separate reduction at a subsidiary it did not name, and the company now expects to be substantially done in Q1 2027. The filing gives a percentage, not a head count.
Here's what changed in the filing:
- Board approval: Nov 10.
- Scope: about 37% of total workforce, up from about 24%, "together with a separate, independent reduction in force initiated by a subsidiary."
- Cost: pre-tax restructuring charges of $27m to $37m, up from $16m to $23m.
- Timing: substantial completion in Q1 2027, pushed back from Q4 2026.
- Missing: no head count, and no name for the subsidiary.
The background matters for anyone planning outreach. Third-quarter revenue fell 8.0% year over year to $185.6 million. CFO Erica Gessert steps down Jan 1. Back on May 7, CEO Hayden Brown told staff in a memo that "Two pizza teams are dead," which reads as a change in org design, not just a budget trim. People who worked in small, autonomous teams are the ones whose structure is going away.
How many people is 37%, really?
On the last disclosed base of about 630 employees (end of 2025), 37% is roughly 233 people, and about 82 of them come from the expansion beyond the May 24%. That base is out of date, though, so read "230" as an estimate that could be wrong in either direction.
There are two reasons the number is shaky:
- The base changed after it was published. Upwork's enterprise arm, Lifted, closed its WorkMotion acquisition on Oct 30, 2026. WorkMotion has entities in 30+ countries. If "total workforce" now includes those staff, 37% means more than 233 people. Upwork hasn't said.
- The trackers already disagree. For the May wave, one tracker counts roughly 145 jobs out of a 600-person headcount and another counts 150. If nobody agrees on the first wave, the combined figure is a guess too.
The honest way to say it is "roughly 230 on the last disclosed base." Here's what the filings and Refolk's index support:
| Group / segment | Figure | Source |
|---|---|---|
| Upwork employees, end-2025 | ~630 | 10-K |
| Internal independent workers, 2025 | ~1,600 (2024: ~2,200) | 10-K |
| Implied 37% cut on the 630 base | ~233 | 630 × 0.37 |
| Expansion beyond the May 24% | ~82 | 630 × 0.13 |
| Internal freelancers per employee | 2.5 (2025) vs 3.7 (2024) | 10-K, derived |
| Engineer profiles with "Upwork" in headline, US vs India | 76 vs 122 | Refolk's index |
Why aren't the 1,600 freelancers counted?
Layoff percentages only count employees. The independent workers Upwork hired through its own marketplace were contractors, so they don't count toward the 37% at all. That bench was about 2.5 times Upwork's headcount in 2025, and it was already shrinking before any layoff was announced.
Look at the two years side by side. In 2024 Upwork had about 600 employees and used about 2,200 independent workers on internal projects. In 2025 it had about 630 employees and about 1,600 independent workers. Staff grew while the contractor bench fell by roughly 27%. Upwork was cutting its own spending on the marketplace before it cut staff.
The order makes sense once you look at what each cut leaves behind. A laid-off employee shows up in a filing, a severance charge and a tracker. A contractor whose work ends shows up in none of them. The quietest cut came first.
A freelancer dropped from Upwork's internal projects gets no WARN notice, no severance line, and no tracker entry.
For recruiters, this group is more useful than the size suggests. These people shipped work for the company that runs the marketplace, they're already set up to work remotely and by contract, and their track record is visible on their marketplace profiles. Nobody is going to send you a list of them, so you have to go and find them.
Which Upwork subsidiary is cutting, and what kind of talent is in it?
The filing doesn't name the subsidiary. The likeliest candidates are inside Lifted, the enterprise arm Upwork launched in 2025, and if it is one of them, the subsidiary cut is mostly a European employer-of-record (EOR) story, not a San Francisco one. Treat that as an inference from Upwork's acquisitions. The filing doesn't confirm it.
Lifted's stack:
- Bubty, based in The Hague. Upwork bought 100% of it on June 27 for $20.4 million in cash.
- Ascen, a YC-backed EOR and compliance company.
- WorkMotion, an EOR based in Berlin. Lifted paid about $129.7 million for it and its entities in 30+ countries. The deal closed Oct 30, 2026.
If the cut lands here, the people affected aren't generalist marketplace engineers. They're contingent-workforce operations people, compliance specialists, and payments engineers who have built for many jurisdictions at once. That profile is rarer than a generalist marketplace engineer, and the two companies with named bases sit in The Hague and Berlin.
These roles are slow to fill. An engineer who has handled multi-country compliance rarely puts that in a job title. Searching for "payments engineer" buries you in fintech generalists, and the few who actually fit are spread across small EOR companies most sourcers have never looked at. This is the kind of specialist search where starting from a pool of already-identified candidates saves the most time.
How do you find laid-off Upwork employees on LinkedIn?
Start with engineer-titled profiles that mention Upwork in the headline, then split the results by country, because the word "Upwork" means something different on a US profile than on an Indian one. In Refolk's index, 76 US engineer profiles have "Upwork" in the headline, compared with 122 in India.
The bigger India number doesn't mean Upwork has more engineers there. In a US sample of 25, 20 list Upwork as their current employer, which is 80%. In an Indian sample of 25, only 8 do (32%). The rest list Cognizant, EPAM, Capgemini, LTIMindtree or EY, and mention Upwork as a side channel for freelance work. The samples are small, so treat those percentages as directional. The pattern is still clear: a Boolean search on "current company = Upwork" returns laid-off staff mixed with freelancers who list the marketplace as their employer.
These signals separate the two:
| Signal | Likely Upwork staff | Likely freelancer |
|---|---|---|
| Title | PM, Engineering Manager | "Top Rated Freelancer" |
| Tenure | Continuous, multi-year | Overlapping gigs, short stints |
| Links | No marketplace profile link | upwork.com/freelancers profile |
| Co-listed employer | None, or prior tech companies | IT-services firm or Fiverr |
Work through it in this order:
- Pull every engineer-titled profile with Upwork as current or recent employer.
- Drop profiles that link to an upwork.com/freelancers page or have a freelancer badge in the title. Keep those for the next section.
- Drop profiles whose main employer is an IT-services firm.
- Rank what's left by tenure and by whether they held a team-lead role. Small-team leads are the people Brown's memo was about.
- Check GitHub for profiles that list Upwork as their company, which catches engineers whose LinkedIn is out of date.
Steps two and three are where most of the time goes. A manual Boolean pass can't easily tell an EPAM engineer who freelances on weekends from someone who spent four years on an Upwork product team. Refolk handles that split: you describe the person you want in plain English, including who to leave out, and get back a ranked shortlist drawn from LinkedIn, GitHub and the open web.
How do you source freelance engineers from Upwork's marketplace?
Go to the marketplace itself, because LinkedIn shows you almost none of these people. Only 88 engineer and full-stack profiles in Refolk's index have "Upwork Top Rated" in their headline, which is a small public footprint next to a 1,600-person internal bench.
The geography of those 88 is narrow. Pakistan leads (Lahore, Islamabad, Karachi, Multan), followed by Dhaka and Lagos. Fiverr often appears as a second employer. Freelancers advertise themselves where their clients look, and that's the marketplace profile, not a LinkedIn headline.
The marketplace has data LinkedIn doesn't:
- Job success score: a measure of client outcomes across past contracts.
- Earnings tiers: a rough sense of how much work the person has been trusted with.
- Verified client feedback: reviews attached to real, paid contracts.
None of those signals tells you which freelancers worked on Upwork's own projects, so you have to ask. A useful opening line is a direct question: "Did you work on any of Upwork's internal projects in 2024 or 2025?" Anyone who says yes has worked inside the company that sets the marketplace's standards.
Most of this group already works remotely, so you're competing with every remote engineering opening they can see.
What should you say to each group, and when?
Base the pitch on what each group lost. Staff lost a salary and their team. Freelancers lost a client. Subsidiary staff in Europe may be losing a specialist EOR job they won't easily replace locally. Because the work runs through Q1 2027, some of these people are still employed and you can reach them before they start a job search.
For the roughly 230 staff: Brown's "Two pizza teams are dead" line tells you what changed. Engineers who liked owning a small team's roadmap now face consolidation. Offer ownership and a named team. For anyone still serving out notice, offer a start date in 2027.
For the subsidiary's team: Be specific about the domain. Mention multi-country compliance or EOR infrastructure in your first sentence, and say where the role is based. Someone in Berlin or The Hague who knows EOR work well won't answer a generic "exciting fintech" message.
For the freelancers: Offer contract-to-hire or a paid trial project, since that's how they already work. Mention their job success score or a specific client review, which shows you actually read the profile. Don't ask them to fill in an application form.
Whatever the group, avoid putting an exact layoff number in your message. The filing doesn't give one, and a candidate who knows the real figure will notice.
FAQ
How many people is Upwork laying off in 2026?
Upwork hasn't given a head count, only a percentage: about 37% of total workforce, up from the 24% previously disclosed. On the end-2025 base of about 630 employees that's roughly 233 people. The WorkMotion acquisition closed Oct 30, 2026, so the real base is probably larger, and the trackers already disagree about the May wave (145 vs 150).
Which Upwork subsidiary is part of the restructuring?
The amended 8-K describes a "separate, independent reduction in force initiated by a subsidiary" but doesn't name it. The likeliest candidates are inside Lifted, Upwork's enterprise arm, which owns Bubty (The Hague), Ascen and WorkMotion (Berlin). That's an inference from the acquisitions, not something the filing confirms.
Are Upwork's 1,600 internal freelancers included in the 37%?
No. The 37% counts employees only. The roughly 1,600 independent workers Upwork used on internal projects in 2025 were contractors, and that bench had already fallen from about 2,200 in 2024. They get no WARN notice, no severance line and no tracker entry, so this group won't show up on any layoff count.
How do I tell an Upwork employee from an Upwork freelancer on LinkedIn?
Check the title, the tenure and the links. Staff tend to have continuous multi-year tenure and titles like PM or Engineering Manager. Freelancers tend to have badges like "Top Rated," overlapping gigs and a link to an upwork.com/freelancers page. In India, most profiles with Upwork in the headline list an IT-services firm such as Cognizant or EPAM as their main employer.
The data behind this
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.
01Describe them
One plain sentence. Role, city, stack, stage, whatever matters to you.
02I read the web live
GitHub, public LinkedIn and Crunchbase records, the open web. Not a database that went stale last quarter.
03You read the shortlist
Ranked, with the reasoning under every name. Open a profile, ask a follow-up, narrow it down.
- Staff backend engineers in NYC who shipped Rust in production
- Series A fintechs in SF under 50 people, growing headcount this year
- Maintainers of fast-growing Rust web frameworks on GitHub
- 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.