From GitHubcommits since 1 MayEngine Room
For business ownersBy the numbers

Google's SEO advice, against 5,599 UK business websites

Google's own advice is short and relaxed. We read it, then checked the homepages of 5,599 UK gyms, tyre fitters, accountants and agencies. Only 27% passed six basic checks together.

27% set large on a pale grey background beside a column chart of the share of 5,599 UK business homepages with a sitemap, a canonical tag, a meta description, exactly one main heading and a tap-to-call link, above the words: Of 5,599 UK business homepages pass six basic checks together
5,599UK business homepages analysed, 12 November 2026
27%pass title, description, one main heading, canonical, sitemap and a real 404 page together
61.1%have a meta description
36.9%have a tap-to-call phone link

Google’s advice on SEO fits on a few pages, and most of it is relaxed. There is “no magical word count target, minimum or maximum”. The keywords meta tag is not used. Heading order does not matter for Search. A sitemap is optional for a site of about 500 pages or fewer, if the pages link to each other. The first tip in the guide is to make content people find useful.

So we asked a plain question: do small UK businesses even do the easy parts? On 12 November we checked the homepages of 5,599 of them.

Who we checked, and how

The sites came from our own lists: gyms and martial arts clubs (3,908), tyre fitters (1,191), web and design agencies (313) and accountants (187). That is every site those lists hold: 7,564 domains, of which 6,039 answered. We dropped 440 pages under 3KB (holding pages and redirect stubs) and kept 5,599.

It is a passive check. One request to each homepage, plus robots.txt, the usual sitemap addresses, a made-up address to test the 404 page, and the www or non-www twin. We read the raw page and ran no scripts, so a title added by JavaScript after load would show as missing. 1,525 sites did not answer, and sites that block automated visits are under-represented. These are four trades, not “all small business”, and we report totals only. No business is named.

The basics are missed more than you would think

Share of the 5,599 homepages that do each thing

Raw homepage only, 12 November 2026.

HTTPS94.1%Sitemap found74.7%Canonical tag67.3%Meta description61.1%Structured data51.6%Exactly one main heading48.7%Tap-to-call phone link36.9%

Source: zxn.uk's own scan, 12 November 2026

For scale, the HTTP Archive’s Web Almanac 2025 crawled 16,213,084 sites. 98.6% have a title and 91.7% use HTTPS, because platforms do those for you. About 67% have a meta description and 68% a canonical tag. Our small businesses sit near the web at large, a little lower, with the same split: what the platform writes automatically is nearly universal, and what a person has to write is where it falls over.

Add the checks together and it gets starker. Only 49.5% of homepages have a title, a meta description and a main heading all at once. Add a canonical tag, a sitemap, exactly one main heading and a proper 404 page, and just 27% pass the lot.

Some specifics from the scan:

The platform matters more than the trade

On WordPress, 96% of sites have a canonical tag and 95% a sitemap. On Squarespace, 98% have a sitemap and 98% have structured data. On the 2,655 sites we could not tie to a common builder, only 38% have a canonical tag, 54% a sitemap and 29% structured data. Webflow sites had a main heading 98% of the time but a canonical tag only 15%.

Where a site was built tells you more about its SEO basics than which trade it is in. If your site came from a developer who set everything up by hand, nothing was set up for you. That is not an argument against custom sites. It is an argument for checking what they left out.

Cheap things almost nobody does

None of these is a ranking switch, and Google has not asked for them. They cost an afternoon.

And one thing to leave alone: FAQ markup is on 4.3% of sites. Google cut back FAQ rich results in 2023, so the low number is the right answer.

The new front door

The Web Almanac found GPTBot named in 4.5% of robots.txt files in 2025, up from 2.9% the year before, and ClaudeBot up from 1.9% to 3.6%. Those count files that mention a crawler, not files that block it. In our scan 12.1% of homepages’ robots.txt named at least one AI crawler: 19.2% of agencies, 13.8% of gyms, 5.9% of accountants. Nobody has to have a policy. But the 88% who have never written one are choosing by default.

Check your own in thirty seconds

We turned the same 18 checks into a free tool. Type your address into the homepage basics check and it shows what passes, what to fix in plain words, and the share of the 5,599 sites that pass each check. Nothing is stored and nobody is emailed. For anything deeper, the free website test and Competitor check on zxn.uk go further. If you have just had a site built, our checklist before a client site goes live and how to tell if it is bringing in enquiries are the next two reads.

Where the numbers come from

  1. Google Search Central, SEO Starter Guide, sitemaps overview, title links, snippets and structured data policies, read 12 November 2026.
  2. Google Search Central blog, rich result changes in June 2025 and November 2025, and the 2023 FAQ and How-to change.
  3. HTTP Archive, Web Almanac 2025, SEO chapter and methodology: 16,213,084 sites, July 2025 crawl. Figures quoted as the chapter states them.
  4. zxn.uk's own passive scan, 12 November 2026: one request per homepage, plus robots.txt, the usual sitemap addresses, a made-up address and the www or non-www twin. 5,599 homepages analysed from 7,564 domains.

Mark Lovelady, Director. Writes a lot of our code and learns from every project.