Last Updated on July 10, 2026
UK small businesses have finally fixed the old website basics. Every site we checked was mobile-friendly and secure. But almost none carry the signals AI search engines use to understand and recommend a business, which means they are increasingly invisible exactly where customers are starting to look.
Key facts
The short version
- The old basics are done: 100% of the 35 sites were mobile-friendly and on HTTPS.
- The AI signals are missing: 91% had no FAQ schema and 37% had no business-type structured data.
- 14% had no structured data at all, so an AI crawler cannot reliably tell what the business even is.
- 9% were missing a meta description and roughly half had no click-to-call link.
- It is achievable: the two best sites in our sample were both independent electricians, with full business schema and FAQ markup. The gap is effort, not budget.
On this page
What we did
We took 35 real UK small business websites, pulled from Google’s local and organic results across a dozen trades and a dozen towns and cities, from electricians in Nottingham to beauty salons in Glasgow and accountants in Leeds. We then checked each homepage for the signals that AI search engines and Google use to understand, trust and recommend a business: structured data, FAQ schema, metadata, mobile-friendliness and contactability.
The pattern was consistent. These businesses have spent the last decade fixing the things the old web rewarded, and most have not started on the things the AI web rewards.
The old basics are fine
Credit where it is due. Every single site was served over HTTPS and carried a mobile viewport, so they render properly on a phone. Nearly all had a page title and most had a meta description and Open Graph tags so they share tidily on social media. A decade of “make sure your site is mobile-friendly and secure” has worked.
The trouble is that those are table stakes now, not advantages. They help a human who already found you. They do little to help a machine decide whether to put you in front of someone who has not.
The AI signals are missing
This is where it falls apart. Structured data is how a site tells a machine, in plain terms, “this is a hair salon in Leeds, here is the address, here are the services, here are the answers to common questions.” It is the difference between a page an AI can confidently summarise and one it has to guess at.
Over a third of the sites had no business-type structured data at all, and 14% had no structured data of any kind. The single biggest gap was FAQ schema: just three sites in 35 had it. That is the format most readily pulled into AI answers, and the small business web has barely touched it.
The full scorecard
| Signal AI and search engines use | Sites passing | Share |
|---|---|---|
| Mobile-friendly (viewport) | 35 / 35 | 100% |
| Secure (HTTPS) | 35 / 35 | 100% |
| Page title | 35 / 35 | 100% |
| Meta description | 32 / 35 | 91% |
| Open Graph (shareable) | 31 / 35 | 89% |
| Any structured data | 30 / 35 | 86% |
| Business / LocalBusiness schema | 22 / 35 | 63% |
| FAQ schema | 3 / 35 | 9% |
| Click-to-call link | 18 / 35 | 51% |
Read top to bottom and the story is clear: the closer a signal is to “old SEO”, the more sites have it; the closer it is to “how AI reads a page”, the fewer do.
Why this matters now
Customers are increasingly starting in an AI answer, not a list of blue links. When someone asks an assistant to recommend a plumber or a salon, the engine has to understand the candidate businesses well enough to name one. A site with no structured data, no FAQ content and little readable text gives it almost nothing to work with. The businesses that are easy for a machine to read are the ones that get named.
It is worth adding that our audit was generous. We checked each site as a browser renders it, with JavaScript run. Many AI crawlers do not run JavaScript, so for the image-heavy and app-style sites in our sample, the real picture an AI sees is likely thinner still.
How to fix it
None of this is expensive. The two best sites in our sample were independent electricians, not national chains, and they had everything: business schema, FAQ schema, a strong page of readable text and a tap-to-call button. The fix for everyone else is the same short list: add LocalBusiness or Organization structured data, add an FAQ section with real questions and matching FAQ schema, write a clear meta description, make sure there is genuine readable text on the page rather than just images, and add a click-to-call link. For the wider context on why AI discovery now matters, see Google AI Overviews 2026 and how AI search has hit UK website traffic.
Cite this research
Whito Research (2026). Invisible to AI: We Audited 35 UK Small Business Websites for AI Readiness. Almost None Are Ready. Whito. https://whito.co.uk/research/uk-small-business-websites-ai-readiness-study/
Key finding: 96 per cent of 23 audited UK SME websites carry no FAQ schema, leaving AI assistants nothing structured to quote.
This is original Whito research. You are welcome to reuse these figures with a link to this page as the source.
Common questions
Are UK small business websites ready for AI search?
Mostly not. In our audit of 35 UK small business websites, 91% had no FAQ schema, 37% had no business-type structured data, and about half had no click-to-call link. They have the old basics (every site was mobile-friendly and on HTTPS) but lack the signals AI engines use to understand and recommend them.
What is the most common AI-readiness failure on small business sites?
Missing FAQ schema. Only 3 of the 35 sites we checked had it. FAQ structured data is one of the clearest ways to feed AI engines question-and-answer content they can lift and cite, and almost no small business is using it.
What should a small business add to get found by AI?
Structured data that names the business and what it does (LocalBusiness or Organization schema), FAQ schema with real questions and answers, a clear meta description, visible text content rather than image-only pages, and a click-to-call link. Most sites we checked had the mobile and security basics but few of the AI-specific signals.
Methodology. We audited 35 UK small business websites in June 2026, selected from Google local and organic results across a dozen trades (including electrician, plumber, hairdresser, accountant, cafe, cleaning, car servicing, florist, beauty, dentist, builder, roofer and personal trainer) and a dozen UK towns and cities. National chains, directories and platforms were excluded. Each homepage was checked as rendered in a browser for: HTTPS, mobile viewport, title, meta description, Open Graph tags, JSON-LD structured data and its types, FAQ schema, an H1 and a click-to-call link. This is an indicative snapshot, not a statistically representative census of the millions of UK small business websites. Because pages were checked with JavaScript rendered, sites relying on client-side rendering may appear better here than they do to AI crawlers that do not run JavaScript. This is a Whito research piece, not advice about any specific business.
Sources
- Whito original audit of 35 UK small business websites, June 2026
- Schema.org structured data vocabulary, schema.org
- Whito research: Google AI Overviews 2026, and AI search and UK website traffic
