“WordPress powers a huge share of the web” is one of the most repeated lines in the industry — and on its own, one of the least useful. Knowing that a plurality of sites run WordPress tells you nothing about which sites, where they are hosted, what version they run, or what else is installed alongside the CMS. Those are the questions that matter if you build plugins, run an agency, do outreach, or maintain sites for a living.
So rather than repeat the “40% of the internet” headline, this article looks at the specific population: the websites made by WordPress that are live and responding right now. The numbers below come from webatla’s WordPress technology dataset, which fingerprints the CMS across the domains it indexes and re-exports the full list daily. As of the latest export, that is 22,333,009 live WordPress sites — about 7.3% of every domain in the index.
One clarification before the data. This counts sites that currently respond and expose a WordPress fingerprint — not parked domains, not installs hidden behind aggressive firewalls, not every site that has ever existed. Read it as a floor for “reachable, detectable WordPress,” not a full census. With that framing, here is what the population actually looks like.
How many websites are made by WordPress?
The headline figure people quote — WordPress running more than 40% of all websites — and webatla’s 7.3% are not in conflict. They use different denominators. The 40%+ figure measures share among sites that run a known CMS. The 7.3% is share of a much broader domain index that also includes millions of parked domains, redirects, and infrastructure hosts with no CMS at all. Both can be true at once.
The practical point survives either way: WordPress is not a niche. It is the single most common site-building technology by a wide margin, and 22.3 million reachable installs is a large, addressable population — not a rounding error.
Where websites made by WordPress are hosted
Grouped by hosting country, the top five account for more than two-thirds of detected sites:
| Country | WordPress sites | Share |
|---|---|---|
| United States | 9,016,743 | 40.9% |
| Germany | 3,150,773 | 14.3% |
| France | 1,136,779 | 5.2% |
| Japan | 979,509 | 4.4% |
| Netherlands | 696,275 | 3.2% |
The US leads, but not overwhelmingly — Germany alone holds 14.3%, and the long tail runs across the EU and Japan. WordPress is genuinely global. One caveat: this is grouped by hosting IP, which is not always the same as where the business or audience sits. A German-hosted site can serve a US audience and vice versa.
Which TLDs dominate
| TLD | WordPress sites | Share |
|---|---|---|
| .com | 10,496,827 | 47.0% |
| .de | 1,201,417 | 5.4% |
| .org | 911,975 | 4.1% |
| .uk | 646,317 | 2.9% |
| .br | 600,365 | 2.7% |
Notably, .com accounts for under half of all WordPress sites — meaning the majority sit on country-code and other generic TLDs. Germany’s .de and Brazil’s .br punch above their weight, which lines up with the country data and confirms Brazil as one of the largest WordPress markets, even though it did not crack the top five by hosting country.
Version spread: where the maintenance risk lives
Among sites where a version could be detected, the distribution is top-heavy but far from uniform:
| Version | WordPress sites | Share |
|---|---|---|
| 6.9.4 | 7,808,735 | 51.2% |
| 7.0 | 2,464,192 | 16.1% |
| 6.8.5 | 1,039,199 | 6.8% |
| 6.7.5 | 388,336 | 2.5% |
| 6.6.5 | 194,018 | 1.3% |
Roughly half of detected sites run the current 6.9.x branch, and 16% have already moved to 7.0. The story, though, is the tail: millions of sites still sitting on 6.6, 6.7 and 6.8 releases. That tail is where unpatched vulnerabilities and abandoned installs accumulate. For anyone doing security or maintenance work, “which version is this site on” is usually a more actionable question than “is it WordPress” — and it is answerable at scale only if the version is attached to each domain.
What runs alongside WordPress
The stack around WordPress is as revealing as the CMS itself. The technologies most frequently detected on the same sites:
| Technology | Type | Sites |
|---|---|---|
| LiteSpeed Cache | Caching plugin | 3.7M |
| Hello Elementor | Elementor theme | 2.0M |
| Astra | Theme | 2.0M |
| Slider Revolution | Plugin | 1.9M |
Caching layers, page-builder themes and specific plugins tell you how a site was built and, often, by whom. For a plugin or theme maker, this is the competitive map: how many sites already run Astra, how many are on a competing builder, and where the realistic switching opportunities are.
Why the list beats the percentage
Every use case for this data has the same shape — it needs the underlying list, not the headline:
- Agencies and freelancers: build targeted prospect lists — WordPress sites in a specific country, on an outdated version, or running a plugin you can replace.
- Plugin and theme developers: size a market, track adoption over time, and find sites currently running a competitor.
- Security and maintenance providers: flag installs on old branches at scale instead of one site at a time.
- Analysts and researchers: measure real CMS adoption from data rather than requoting the same recycled percentage.
None of that is possible with a single “40% of the web” statistic. A percentage cannot be filtered, sorted, or exported. A structured dataset — one row per site, each carrying the WordPress version, hosting country, TLD and co-installed technologies — can be queried against any of those criteria.
Getting the underlying data
webatla exports the full population of websites made by WordPress as JSON Lines, refreshed daily, with a free sample so you can inspect the schema before committing to anything. Each row pairs the domain with its detected technology stack, IP and hosting country, TLD and ranking — which is what turns “websites made by WordPress” from a talking point into something you can actually query.
The web’s most common CMS deserves more than a rounded-up percentage. 22.3 million live installs is a population you can study properly — as long as you can see it one row at a time.