WordPress web share falls from 43.0% to 38.7% in 2026, breaking a decade of growth

A decade of growth comes to an end

According to figures from W3Techs cited by developer Hannan Shoukat in a dev.to post published on 14 November 2026, WordPress powered 43.0% of all websites on 1 January 2026. By the time the post was written, that figure had fallen to 38.7% — a drop of 4.3 percentage points in under a year. The decline breaks a long run: Shoukat notes that WordPress had grown almost without interruption for a decade, rising from 23.3% of the web in 2015.

The post uses the number as a prompt rather than an obituary. If a small business asked for a new website today, the author asks, would he still choose WordPress?

Where the lost share went

The most obvious competitors did not absorb the losses. According to the post, Shopify, Wix and Squarespace did not account for most of the gap. Instead, the decline shows up largely in sites that W3Techs cannot attribute to any CMS at all.

Shoukat's own interpretation — which he explicitly flags as possibly wrong — is that these are custom-built sites and AI-generated sites. That reading should be treated as a hypothesis rather than a measurement: a "no CMS detected" bucket covers everything a crawler cannot classify, and it does not prove how those sites were made.

The case for staying with WordPress

Shoukat still defaults to WordPress in most client situations, and his reasons are practical rather than ideological. Business owners can edit their own pages without calling a developer, which he argues matters to them more than any technical consideration. Hosting is inexpensive and available from many providers, so no single company can hold a client hostage on price. The plugin ecosystem means common requirements — e-commerce through WooCommerce, forms, SEO tooling — are already solved. And if the original developer disappears, another one can take over the site without a steep learning curve.

The reasons for hesitation

The author is less comfortable than he used to be, for four reasons. A deceptively simple site can accumulate twenty-five plugins, each one a potential point of failure or a route in for attackers. Software updates remain the client's responsibility until an emergency turns them back into the developer's problem. For a five-page brochure site, a static site or a hosted builder is faster and demands almost no maintenance. And page builders, while they make editing easy, also make it easy to produce slow websites.

A working rule of thumb

Shoukat's current approach splits the decision by use case. If the client will update content often or sell through the site, WordPress remains his choice. If the site barely changes, he picks something lighter. He concedes he keeps changing his mind about the cases in between, and closes the post by asking other developers what they have moved clients to, and whether they regretted it.

Why it matters

A 4.3-point decline still leaves WordPress far larger than any rival — 38.7% of all websites is dominance most platforms can only imagine. But the direction is new, and the composition of the loss is arguably more interesting than its size. If the unattributed sites really are custom builds or AI-generated output, WordPress is not losing to a competing CMS but to a shift in how websites get made in the first place.

For agencies and freelancers, the practical lesson is portfolio-level. Defaulting to one CMS for every client no longer reflects the market. Choosing the lightest tool that fits the job — and keeping plugin counts deliberately low when WordPress is the right choice — has moved from good hygiene to a first-order business decision.