There is a particular kind of layoff that is more interesting than the headline number. It is the layoff where the company being shrunk built its original business on automating away someone else’s job, and is now being automated away by the next layer up.
On May 25, 2026, Wix.com Ltd announced it will cut roughly 1,000 jobs — about 20% of its headcount — in what CTech called the “largest layoff round in company history.” It is the first major publicly listed SaaS company in this cycle to put the AI-disruption explanation directly into the layoff press release rather than route it through the standard “organisational streamlining” euphemism that Intuit’s CEO used the week before.
What the numbers say
Wix ended Q1 with 5,277 employees, more than 60% based in Israel. The 1,000-headcount reduction lands at roughly 19% of that base. Q1 results explain the urgency:
- Revenue up 14% year-on-year to $541M. A respectable growth print.
- Net loss of $57.5M. After several consecutive profitable quarters, Wix returned to losses.
- Operating expenses up 50% to $423M — from 21% of revenue in Q1 2025 to 35% of revenue in Q1 2026.
- Cash flow down 21% to $112M.
- Stock down ~50% year-to-date, market cap roughly $2B.
- $1.6B buyback executed in March, intended to restore investor confidence, drained cash reserves to ~$900M and so far has not moved the share price.
The Q1 headcount line is the most interesting one. Before May 25, Wix had reduced its workforce by 63 employees in the first quarter — a rounding error. The 1,000-cut announcement is what happens when a company that was trying to grow through the disruption discovers that the growth is more expensive than the disruption and has to switch modes mid-flight.
What the strategic position says
Wix’s founding pitch, since 2006, was a website builder for people who did not want to hire a web developer. That pitch competed against a labour category — freelance HTML/CSS/PHP web developers — and it won. The market shifted. The freelance web dev was disintermediated by drag-and-drop. Wix sold the picks and shovels.
The thing happening now is the same trade compressed by a generation. The next disintermediation does not go through a drag-and-drop UI. It goes through an LLM coding tool. You type “build me a portfolio site for a wedding photographer with a booking form” into Lovable, Bolt, V0, Cursor, or the dozen new vibe-coding tools shipped this quarter, and the output is a working site. The user no longer needs a website builder as a category of product. The product is the LLM.
Wix saw this coming. In 2024 it acquired Base44, the natural-language coding platform founded by Maor Shlomo, and slotted it into the Wix stack as the AI-native answer to the next-generation tooling. The acquisition is working. According to today’s press release, Base44 reached $150M annual recurring revenue in May, “significantly ahead of internal targets,” and is responsible for most of Wix’s current growth. Wix aired two Super Bowl ads this year — one for its core products, one for Base44 — which tells you which property they are leaning on.
The structural problem is that Base44 working is exactly what undermines Wix working. The faster Base44 grows, the faster Wix’s traditional templated-builder roles — front-end engineers maintaining a template library, support reps walking small-business customers through a drag-and-drop UI, sales staff selling a SKU that an AI prompt is now replacing — become surplus. Defensive M&A has accelerated the cannibalisation, not slowed it.
There is also a direct cash cost. Base44’s earn-out structure paid Shlomo $38M last quarter alone, and Wix warned investors that additional payments will be required later this year as Base44 hits successive ARR milestones. The faster Base44 grows, the more cash leaves Wix to the founder of the thing eating its base business. The opex line gets hit twice: once by the Base44 marketing spend, once by the earn-out.
The third hit is compute. A drag-and-drop builder is cheap to run: the user assembles a template on the client side, the server stores a row in a database. A vibe-coding tool runs an LLM inference for every meaningful interaction. Inference at scale is the new line item in Wix’s P&L, and the company told investors it is building its own AI model for Harmony — Wix’s AI website-building system — in part because the long-run unit economics of paying a frontier-model API per prompt are not workable at the price points Wix charges.
What Abrahami is not saying out loud
CEO Avishai Abrahami’s investor letter framed the model build as a path to lower inference costs and better Harmony accuracy. Read at face value, that is a roadmap. Read with the May 25 cut sheet next to it, it is the same admission Intuit made the week before and Microsoft made back in April: the company has to spend more on compute and AI development to stay relevant, and the way it pays for that spend is by spending less on people.
The pattern is now boringly consistent across the May 2026 cohort. Meta cuts 8,000 to fund AI (May 20). Intuit cuts 3,000 to fund AI and explicitly denies it is for AI (May 20). Cloudflare cuts 1,100 and credits a 600% internal AI usage spike (May 20). Workday turns a $500M AI ARR into a 14M-recruiter agent and an HCM platform (May 21). Wix cuts 1,000 to fund AI and at least says so plainly.
The “at least says so plainly” part matters. The companies that name AI as the cause give investors something to grade them on; the companies that route around the explanation get punished twice, once by markets that assume worse, once by employees who can read the org chart. Wix has chosen the first lane.
What to watch
- Israeli tech-sector AI layoff print. Wix is the largest of these so far, but Globes and Ynet reporting suggest a wave of mid-cap Israeli tech firms in the same SaaS-margin squeeze. Watch monthly tech-sector layoff prints out of the Israel Innovation Authority for follow-on cuts.
- Base44 second-year ARR. Wix has told the market Base44 hit $150M ARR ahead of plan. The Q3 number is the one that decides whether Wix is the company that successfully reinvented itself through M&A, or the company that paid $38M-per-quarter earn-outs to a founder and watched its core business shrink faster than the new one grew.
- Wix’s own AI model. The Harmony in-house model has the structural fingerprint of an AWS Bedrock / Vertex / Azure OpenAI cost-control project. The interesting tell is whether Wix discloses a per-prompt unit cost trajectory; if it does, that is the line item against which every other AI-built-on-someone-else’s-model SaaS company will be benchmarked.
- The Klarna outcome. Wix has cut explicitly to fund AI investment. The Klarna case — 700 customer-service jobs cut, replaced by AI, then rehired after quality collapsed — is the canonical cautionary tale. If Wix’s small-business customers start churning to competitors that still offer human onboarding support, the next press release writes itself.