Where the Upwork Work Went: The 2026 Tool Rotation Data

WordPress is bleeding 2,162 Upwork jobs per month against where it was a year ago. The five biggest "new" tools combined — Lovable, Cursor, Shopify, Squarespace, Replit — picked up 1,205 per month. The substitution that everyone is writing about is real, but it is incomplete, and the gap is the part nobody is putting on a chart.

This post is that chart, plus what it actually means for freelancers and agencies positioning on Upwork in 2026.

The numbers in one table

Sorted by monthly job change, April 2026 versus January 2025:

| Tool | Jobs/month change | Direction | | --- | --- | --- | | WordPress | -2,162 | falling | | Wix | -159 | falling | | Webflow | -137 | falling | | jQuery | -92 | falling | | Joomla | -44 | falling | | Net legacy loss | -2,594 | | | Replit | +59 | rising | | Squarespace | +96 | rising | | Cursor | +295 | rising | | Lovable | +370 | rising | | Shopify | +385 | rising | | Net new gain | +1,205 | | | Net across ten tools | -1,389 | contracting |

WordPress alone sheds more jobs per month than the five biggest gainers combined replace.

What the chart is and is not saying

It is saying: the Upwork "build me a website" category, broadly defined, is contracting in 2026. The legacy stack is dying faster than the new stack is being adopted. The narrative of "WordPress is being replaced by Lovable" is partially true and meaningfully incomplete.

It is not saying that Upwork is shrinking everywhere. Other categories — AI development, automation, specialized engineering — are growing. Total Upwork volume is down about 20% over the same window (we covered that separately), but a 20% top-line drop is not enough to explain WordPress's 27% drop.

The category is contracting on three axes at once:

1. Some of the work is being absorbed by the new tools. Lovable and Cursor are real businesses doing real work. A non-trivial slice of the 2,162 WordPress jobs/month that disappeared have moved to "I want a Lovable site" or "Cursor agency wanted."

2. Some of the work is being absorbed in-house by AI. A site that took a freelancer a week in 2024 takes a non-technical founder an afternoon with Lovable in 2026. That work doesn't move to a different freelancer. It stops being freelance work.

3. Some of the buyers moved off Upwork entirely. Shopify's +385/month suggests one segment is still hiring help — but they're hiring it for Shopify store optimization, not for WordPress sites. The intent shifted, the platform did too.

Where this leaves each side of the market

If you market yourself as a WordPress developer

Your buyer pool is contracting. Not changing nouns, contracting. The honest read is that "WordPress developer" as a positioning is going to keep losing budget share each quarter for the foreseeable future.

Three responses make sense:

• Pick a second tool. Lovable, Webflow, or Shopify all have growing or stable demand. Adding a second specialization isn't a personality change — it's keeping your portfolio current.
• Move up the stack. "WordPress developer" is replaceable. "WordPress migration consultant" or "WooCommerce performance engineer" addresses problems that AI tools can't solve in an afternoon, and the budget pool is more durable.
• Pick a second platform. Upwork is not the only place clients hire. The clients who still want full WordPress builds are increasingly going to dedicated agencies, niche marketplaces, or direct hire — channels where the "I need a WordPress dev" search is the entry point.

If you market yourself as a vibe-coding specialist (Lovable, Cursor, etc.)

You are in the right category, but watch the absolute scale. Lovable is up 627%, which is a great relative number, but at 429 jobs in April 2026 it is still 14 times smaller than WordPress. The market is not winner-take-all yet, and it will not be for at least another year.

The realistic take: build your portfolio around the new tools while keeping enough range to do classical work when it pays. The "Lovable specialist" positioning is genuinely promising, but it should be one of two specializations, not the only one.

If you run an agency on Upwork

The agency play in this market is straightforward: be the bridge. The clients who have a Lovable demo and want a "real" site are the same clients who need WordPress migration help, custom integrations, multi-platform deployments. Position the agency as the team that handles the AI-tool-to-production gap, not as another single-tool shop.

Why this isn't just a WordPress story

WordPress is the loudest line on the chart because it is the largest in absolute terms. But the same pattern shows up in adjacent legacy categories:

• jQuery is down 56% — the largest percentage decline in the dataset
• Joomla is down 54%
• Drupal is down 41%

The whole "old-web" stack is bleeding at double-digit percentages. The vibe-coding tools are not catching all of it. Some of the work is being absorbed in-house, some is leaving Upwork, and some is just not getting done at all because AI raised the bar on what counts as worth hiring out.

The contrarian read: this is not a tools story. It is a category-of-work story. "Build me a basic website" is becoming non-freelance work the way "design me a logo" became non-freelance work after Canva. The freelancers who survive this transition are the ones doing work that AI can't yet do in an afternoon — integrations, migrations, performance, security, custom backends, specialized industries.

Where this data comes from

UpHunt scrapes every public Upwork job posting and stores it in a queryable database. The dataset is 3.6 million+ jobs across two-plus years, and grows by roughly 5,000 new postings a day. This analysis covers the full sixteen-month window from January 2025 to April 2026 — 2,974,845 jobs in total.

The same dataset powers UpHunt's AI proposal automation. Every new Upwork job gets matched against your profile, scored for fit, and (for paying users) auto-applied to with a proposal written by Claude that references the client's actual context. If you want to act on this market data instead of just reading about it, start with UpHunt — it is the same engine, pointed at your outreach instead of your blog reading.

The short version

The category is contracting, not rotating. WordPress is bleeding. Lovable and Cursor are real but tiny by comparison. The five biggest new tools combined replace less than half of what WordPress alone sheds each month.

If you're a WordPress-only developer, add a second tool. If you're a vibe-coding-only specialist, add a second tool. The Upwork market in 2026 rewards range plus depth, not single-tool monogamy in either direction.

We re-run these queries monthly. The next major shift in the data will land here first. Subscribe to the blog if you want the next chart in your inbox.