I Stopped Recommending WordPress to New Clients. Here's My Exact Cutoff.

Somewhere in early 2026 I noticed I had stopped writing "WordPress" in my proposals unless a client asked for it by name.

Not as a policy. Nobody sat down and decided anything. I just kept getting to the stack line in the scope document and typing something else, and one day I scrolled back through the year and realized the pattern was already there.

So I went and wrote the rule down, because an unspoken default is a bad default. Here it is: WordPress goes on the table when the site needs a piece of software running behind it that WordPress already solved. If it only needs pages, it does not go on the table. That is the whole cutoff. The rest of this post is what that means in practice, where I still recommend WordPress without hesitating, and how you can apply the same test to a site you already have.

The question people actually ask me is "should I still use WordPress in 2026?", and the honest answer is that it turns on exactly one thing, and it is not performance.

This is not a hit piece. I built on WordPress for years, I have written plugins for it, and I would still hand it to some of the clients who walk in tomorrow.

The Number That Made Me Check My Own Default

WordPress peaked at 43.6% of all websites in January 2025. It was at 43.20% in December 2025. By late May 2026 it was 41.90%, six consecutive months of decline, and Search Engine Journal noted at the time that the slide had accelerated from a modest 0.60% drop the year before. When I checked W3Techs on the day I wrote this, it read 40.7% of all websites.

Now, hold that next to the other number on the same page: 58.9% of all sites whose CMS is known still run WordPress. Not a collapse. A platform coming off a historic high, still roughly three times the size of everything else combined, losing a percentage point or so a year to Wix, Shopify, Squarespace and Webflow.

If you look at that graph and conclude WordPress is dying, you are reading it wrong. What I read into it is narrower and more useful: the sites leaving are mostly small. Nobody migrates a 900-product WooCommerce store to Squarespace. Brochure sites and small service businesses are the ones that leave, because those are the sites where WordPress was always the most expensive answer to the simplest question.

That is the same population I was quietly steering elsewhere. The market and my proposals had drifted in the same direction, and I had not thought hard about why. So I did.

My Actual Cutoff, in Three Questions

When a new inquiry comes in, I ask three things before the stack even gets discussed.

One: is there software, or are there only pages? A checkout, a membership wall, a booking calendar with availability rules, a directory with faceted search, a multi-language editorial workflow. That is software. A homepage, an about page, six service pages, a contact form and a blog is not software. It is a document with a nice layout.

Two: how many people touch the content, and how often? One marketing manager updating a page twice a month is a completely different problem from six people publishing daily with an editor, a proofreader and a legal reviewer in the chain.

Three: what already exists that nobody wants to rebuild? If the team has spent four years building forms, automations and CRM syncs on top of a specific plugin stack, the cost of leaving is not the website. It is the workflow.

If the answers are "software", "many people", or "years of plugin-shaped process", WordPress is on the table and often wins outright. If the answers are "pages", "one or two people", and "nothing", it is now genuinely hard for me to defend, and I will explain why in a minute.

Note what is not in those three questions: my personal preference, performance benchmarks, and whether the client thinks WordPress sounds outdated. Those come up constantly and none of them should decide anything. If you are asking should I still use WordPress in 2026 because a competitor's site loads faster, you are asking about a symptom. I wrote about the framework I use once WordPress is on the table in WordPress, Next.js, or Astro? Here's How I Actually Decide for Client Projects. This post is the step before that one: whether it belongs in the comparison at all.

Where WordPress Still Clearly Wins, and I Say So

Start with commerce, because it is the least ambiguous case. WooCommerce runs somewhere around 4.15 million live stores and roughly a third of all online stores by store count. If a client is selling physical products at small to mid volume in a market with tax and shipping rules that a plugin already handles, "let's build a custom checkout" is not an idea, it is a bill. Say yes to WooCommerce and move on.

Then the multi-editor newsroom. This is the case people underrate hardest. WordPress's roles and capabilities system, revisions, scheduled publishing, editorial notes and the whole permissions layer are twenty years of accumulated answers to problems that show up on day three of running an actual publication. I have watched teams try to reproduce "the intern can draft but not publish, and legal has to sign off on anything tagged Compliance" in a headless setup, and it takes weeks to get to something worse. WordPress 7.1 in August 2026 added richer editorial notes and shareable links to specific revisions. It is still being built for this exact user.

Third, a non-technical team of six publishing daily. Not because the editor is delightful, but because the training cost is near zero. Somebody on that team has used WordPress before. Somebody's cousin can cover when the marketing manager is on holiday. That is a real operational asset and it does not appear in any performance benchmark.

Fourth, plugin-dependent workflows nobody wants to rebuild. If Gravity Forms feeds a CRM which triggers a membership tier which gates a resource library, and it has worked for four years, the correct professional advice is often "leave it alone, harden it, and spend the budget on something that moves revenue." I have given that advice and lost the project. That was the right outcome.

One more thing worth saying plainly: the project is not abandoned. WordPress 7.0 shipped in May 2026 with more than 875 contributors and translations in 70-plus locales. Whatever else is true, that is not a dead platform.

Where It Is Now Hard for Me to Defend

The brochure site. The 20-page service business. The content site with one or two editors. These are the ones where I have stopped reaching for WordPress, and the reason is maintenance surface, not speed.

Patchstack's State of WordPress Security in 2026 counted 11,334 new vulnerabilities in the ecosystem during 2025, a 42% increase over 2024. 91% of them were in plugins, 9% in themes, and six were in WordPress core, all low priority. High-severity findings rose 113% year over year. And 46% of vulnerabilities did not get a patch from the developer before public disclosure.

Read those numbers carefully, because they do not say what people usually use them to say. WordPress core is not the problem. Core had six low-priority issues in a year. The problem is that a normal small-business WordPress site carries fifteen to twenty-five third-party plugins, each one an independent vendor with its own release discipline, and roughly half of the reported issues sit unpatched at disclosure. The volume is a function of ecosystem size and a genuinely good bug bounty program, which I got into in WordPress Got 773 Security Reports in One Month. But a 20-page plumbing company site does not care about the reason. It carries the surface either way.

So the honest question for a brochure site is: what do those plugins buy you? A contact form. An SEO title field. A caching layer. A cookie banner. Backups, because there is a database to back up. Security scanning, because there is an admin login to attack. Four of those six exist to solve problems the platform itself introduced.

That is the shape of it. On a small site, a meaningful share of the WordPress stack is there to manage WordPress. Once I started counting it that way, I could not un-see it. I have also written up the plugins I think are pure overhead in 20 WordPress Plugins You Can Replace With a Few Lines of PHP, and that list applies whether or not you ever leave.

The Roadmap Is Not Being Written for Your 20-Page Site

Look at what WordPress core shipped in 2026 and ask who it is for.

WordPress 7.0 in May brought a provider-agnostic AI client, an Abilities API, a Connectors screen, and the first admin redesign since 2013. Real-time collaboration, the feature that would have mattered most to that newsroom, was pulled on May 8 over stability concerns and is still not in 7.1. August's 7.1 added interactive state styling, a Tabs block, a Playlist block, and moved image resizing into the browser.

Some of that is genuinely good work. Client-side image resizing fixes something that has annoyed people for fifteen years. But notice the direction: this is a platform investing in being an application framework with an AI layer. If you run a media operation or a plugin business, that is your roadmap and it is being built for you. If you run a dentist's office website, none of it lands, and you inherit a bigger admin surface anyway. I was blunter about that priority ordering in Dear WordPress: Nobody Asked for AI Connectors. We Asked for Caching.

There is also the governance situation, which I will keep short because it is genuinely hard to advise on. The WP Engine litigation is still running through 2026, WP Engine filed a third amended complaint in February 2026, and a motion to intervene has raised WordPress.org's governance structure directly. I do not tell clients WordPress is unsafe because of a lawsuit. That would be dishonest. What I do say is that WordPress.org is not governed like the infrastructure that 40% of the web depends on, and when a client asks me who is accountable for the plugin repository their business runs on, I do not have a clean answer.

So, Should I Still Use WordPress in 2026?

If you already have a WordPress site and it works, the answer is almost always yes, keep it. Migration has a real cost, and "my developer says WordPress is old" is not a business reason.

Here is the test I would run on your own site instead. Open your plugins page and put every active plugin into one of two buckets. Bucket A: this plugin does something a visitor or a customer would notice if it vanished. Booking, checkout, memberships, the calculator on your services page. Bucket B: this plugin exists to make WordPress work properly. Caching, security scanning, backups, an SEO title field, a form, a page builder.

Count both.

If bucket A has three or more things in it and your team is more than two people, stay. Your money is buying you real software and the alternatives will cost more to build than they save. Ask me about hardening and hosting instead, not about migrating.

If bucket A is empty or has one item, and bucket B has ten, you are paying the operating cost of a content management system to serve pages that have not changed in eight months. That is the situation where I would not start a new project on WordPress today, and it is roughly the profile of the three client sites I have moved to Astro. Nothing broke, which I wrote up in I Migrated 3 Client Sites From WordPress to Astro, but I want to be careful not to generalize from three projects into a law of nature. Three is three.

The in-between case, one thing in bucket A and a small team, is genuinely ambiguous and I do not have a clean rule for it. Usually it comes down to whether that one thing has a decent hosted alternative. Sometimes it does. Sometimes the honest answer is that moving would cost more than the annoyance is worth, and I say that.

What I Am Still Unsure About

The part I keep turning over is whether I am pattern-matching on my own convenience. I enjoy building static sites more than I enjoy debugging a plugin conflict at 11pm. That is a preference, not a decision rule, and preferences are very good at dressing themselves up as principles. The three questions at the top exist mostly to keep me honest with myself.

I also do not know what the market share line does next. It could be a normal correction after a fifteen-year run, and 40% of the web is an enormous floor to fall from. Anyone telling you confidently where it lands in 2028 is guessing.

What I am reasonably sure about is narrower. WordPress is a superb answer to "I need software behind my website and I do not want to build it." It has become an expensive answer to "I need eight pages on the internet." Those used to be the same recommendation, and for most of the past decade nobody had to think about it. Now you do, which is why I wrote the rule down instead of leaving it as a habit. I also wrote about the personal side of stepping away in Why I Stopped Using WordPress After 6 Years, and that one is about me, not about client work.

Run the two-bucket count on your own site. It takes four minutes and it will tell you more than any "is WordPress dead" article, including this one. And if you count it up and are not sure what the numbers mean for your situation, that is a conversation I am happy to have over at wunderlandmedia.com.