Website SEO Statistics: What 5,000 Homepages Reveal

How the research sweep was run

The sweep scanned one homepage per site, with one attempt each and no retries. KinetixSEO drew 5,000 domains at random with a public seed from Tranco list 262W9, a ranking of popular domains, not a list of every page on the web. The domain list, the seed and its sha256 hash are all public, so anyone can redraw the same sample. The full aggregate results can be downloaded as CSV and JSON from the edition page. Keep the scope in mind before reading the numbers: these are the homepages of popular domains, not random pages and not the long tail of the web.

What happened when the crawler hit all 5,000 sites?

Of 5,000 domains drawn, 3,096 homepages could be scanned. 1,734 could not be reached or read, 149 were blocked by robots.txt, and 21 failed on the scanner's side and were left out, according to the research sweep. Refusals are a finding of their own: 12.6% of the 5,000 drawn sites refused the crawler, 479 with an HTTP 403, 406 or 429 response and 149 through robots.txt. Each share below counts only the sites where that check could run, so the number of sites behind each share differs from check to check. The scanner image for this edition was pinned before two fixes. It didn't retry a bare domain on "www." when the bare domain never answered, and its robots.txt step lacked the fallbacks the scanner has now. Because of that, a few hundred domains were never scanned, and those lean toward neglected sites. So every share below probably looks a little better than the real population does.

How many websites have structured data?

52% of the 3,028 scanned homepages had no JSON-LD structured data, per the research sweep. Structured data is a block of JSON in a page that tells search engines and AI systems what the page is (an article, a product, a local business) instead of leaving them to work it out from the prose. In search it makes a page eligible for rich results such as product ratings and breadcrumbs. For AI answers, machine-readable labels make a page easier to parse and quote accurately. To check your own site in five minutes, open the page source, search for application/ld+json, and confirm the script names a real type. For writing your own, see this structured data SEO guide.

How many websites are missing an XML sitemap?

No XML sitemap was found for 30% of the 2,984 scanned sites, according to the research sweep. A sitemap is a plain XML file that lists a site's pages, so crawlers can find them without relying only on internal links. It matters most on sites with many pages or weak internal linking, where crawlers can miss whole sections. To check in five minutes, open yourdomain.com/sitemap.xml in a browser, or look for a Sitemap: line in /robots.txt. If neither exists, most content management systems can generate one; the XML sitemap guide covers the rest.

What percentage of websites have no meta description?

20% of the 3,028 scanned homepages had no meta description, the research sweep found. A meta description is the short HTML tag that often becomes the snippet under a search result. Without one, Google picks snippet text from the page body, which is often an awkward fragment, and an AI answer engine loses a short, clear summary of the page. To check, view the page source and search for <meta name="description". This on-page SEO checklist walks through writing one step by step.

How many homepages skip alt text or a contact page?

23% of the 2,027 scanned homepages had at least one image without alt text, and 67% of the 3,028 scanned homepages had no privacy, contact or about link, per the research sweep. Alt text is a short description attached to an image in HTML. It matters for accessibility, for image search, and for any system that reads the page as text. A privacy, contact or about link shows who runs the site, a basic trust signal for readers and search engines alike; E-E-A-T covers the wider idea. Both take five minutes to check: look in the source for <img> tags with no alt attribute at all (an empty alt="" is correct for decorative images), then check the header and footer for a working link to one of those three pages.

How many websites have llms.txt or block AI crawlers?

83% of the 3,028 scanned sites had no llms.txt file, and 13% of the 3,024 scanned sites blocked at least one AI crawler in robots.txt, per the research sweep. llms.txt is a proposed plain-text file that gives AI systems a curated map of a site's content; this explainer covers it in full. Both shares are bounds, for the reasons the edition page gives: the llms.txt share is an upper bound and the AI-crawler share is a lower bound. It is not proven that an llms.txt file, or letting AI crawlers in, changes whether a site gets cited in an AI answer. That is why these files come after the basics here: fix the fundamentals first, then treat the AI-specific files as a second layer.

How do you check your own website's SEO?

The free KinetixSEO website check runs the same scan as the sweep on one site, with no account needed. It covers structured data, the sitemap, the meta description, alt text, the trust links and the AI files described above. To check by hand instead, follow the on-page SEO checklist and the structured data glossary entry. Google Search Console shows what Google itself has indexed and flagged.

What these website SEO statistics mean for site owners

The web's first problem is not AI. It is the basics, and more than half the homepages scanned skip at least one of them. These are homepages of popular, established domains, not small or new sites, and the scanning gaps described earlier mean the real picture is somewhat worse than these numbers show. The sweep measured whether each basic was present, not what it does for rankings, so nothing here shows that adding structured data or a sitemap raises rankings by any amount. If you want to know where your own site stands, start with structured data, a sitemap and a meta description, because they are the most commonly missing and the quickest to check. Then add a visible privacy or contact link and alt text on key images. llms.txt and AI-crawler rules can come after those, not instead of them.

Frequently asked questions

What are the most important SEO statistics from this research sweep?

52% of 3,028 scanned homepages had no JSON-LD structured data, 30% of 2,984 had no XML sitemap, and 20% of 3,028 had no meta description, all from the research sweep edition of 25 October 2026. These three basics are the quickest to check and fix, which is why this article starts with them and covers the AI-specific files later.

How many websites have structured data in 2026?

In this sample, 52% of the 3,028 scanned homepages had no JSON-LD structured data, per the same research sweep. The sample covers homepages of popular Tranco-ranked domains, not every page of a site and not the whole web, so smaller or newer sites may look different.

What percentage of websites have no meta description?

20% of the 3,028 scanned homepages had no meta description, according to the research sweep. It is one of the simplest tags to add, and you can check for it in under a minute by viewing a page's HTML source.

How many websites have an llms.txt file?

83% of the 3,028 scanned sites had no llms.txt file. The sweep reports that share as an upper bound, for the reason the edition page gives. It is not proven that the file changes whether AI systems cite a site, so add it after structured data, a sitemap and meta descriptions are in place.

How can I check my website's SEO statistics myself?

Run the sweep's checks on your own homepage: look for JSON-LD in the page source, open /sitemap.xml, confirm a meta description tag exists, look for images with no alt attribute, and find a contact or about link. The free KinetixSEO website check runs all of these on a single URL without an account, the same scan the 5,000-site sweep ran on every site.