Why I stopped taking WordPress work

I want to be clear upfront: WordPress is not bad software. It runs 43% of the web for a reason. This is about fit, not quality.

We took our last WordPress project in Q3 2025. The client was a Melbourne-based law firm—good brief, reasonable budget, clear scope. We delivered it. It performed fine. Then the maintenance requests started.

The maintenance problem

WordPress sites don't sit still. Plugins update. Themes conflict. WooCommerce has its own release cadence. On a well-maintained site, this is manageable. On a site that hasn't been touched in four months—which is most client sites—it's a quarterly emergency.

The real cost isn't the time to update. It's the time to diagnose what the update broke. We were billing two to three hours a month per WordPress site on average. The clients didn't love paying for it. We didn't love doing it.

What changed on the other side

Next.js with a headless CMS (Sanity or Contentful, depending on the client) now delivers a better content editing experience than the WordPress block editor—and the output is faster by a wide margin. A site built on this stack can hit 95+ Lighthouse scores without heroics.

The deploy pipeline is the other part. Vercel or Netlify means: push to git, site updates. No FTP, no cPanel, no PHP version mismatch. A non-technical client can trigger a deployment by merging a pull request.

“Page builders are technical debt with a good demo. They look great in a sales call and cost double to fix eighteen months later.”

When WordPress still makes sense

Large content teams who already live in the WP admin. Sites with 10,000+ posts where migration cost is prohibitive. Clients with existing WordPress developers on retainer. These are real scenarios. We refer them to good WordPress agencies.

For everyone else building a new site in 2026: the modern stack is faster to build, faster to load, and cheaper to maintain. The learning curve for clients is roughly equivalent.