We audited 2,307 UK care homes for AI search. Not one is ready.

When a family opens ChatGPT and asks for "a good care home near me" or "a nursing home in Leeds that takes dementia residents", the assistant does not return a page of ten blue links. It names two or three homes. To be one of them, a website has to clear the top band of our 0 to 100 AI visibility scale, which we call AI-Ready and which starts at 80. We pointed that audit at 2,307 UK care homes. Here is the headline.

Not one home cleared the bar. The highest score in the entire cohort reached just 78.8, still short of the 80 mark. The cohort average sits at 48.3 out of 100, with a median of 49.6. For a decision families make on trust, that is a real exposure: the moment families start asking an assistant instead of a search box, almost every care home in Britain is invisible to the recommendation. One in four homes, 583 of the 2,307, scores below 40, the point at which an assistant can barely use the site at all.

Reachable, but not quotable

This is not the usual story of neglected, broken websites. On the technical measures, care homes are not far off other sectors. Their Technical category averages 71.7 and AI Citability averages 71.2, meaning the pages load, the crawlers get in, the mobile layout works and nothing is obviously broken. Whatever platforms care homes build on, they handle the plumbing reasonably well.

So the door is open. The AI can reach the site. What it cannot do is understand who runs the home or lift a clean, trustworthy answer from it, because the categories that carry the recommendation are exactly where homes are weakest.

The pattern repeats across the whole index. Homes are reachable, then no one adds the small labels that let an assistant read and trust who runs the home, what it offers and what families say. The door is open; the shelves are bare.

Source: SearchScore SAVI UK Care Homes 2026. 2,307 homes scored across 250+ AI search visibility signals. "AI-Ready" is a score of 80 or higher. Category figures are cohort averages.

Outstanding rated, still invisible

Here is the finding that should stop every operator in their tracks. Of the 2,307 homes, 2,196 carry a current CQC rating. If the regulator's verdict shaped what AI recommends, the Outstanding homes would stand clear of the rest. They do not. The regulator's top badge buys under 4 points of AI visibility, and not one home in any rating band is AI-Ready.

Look at the spread. Outstanding homes average 50.0, Requires improvement 46.1, essentially the same. Rating and AI-findability are, on this evidence, unrelated. Not one of the 125 Outstanding-rated homes is AI-Ready. One in five Outstanding homes (24 of 125, 19%) is effectively invisible to AI, and 41% of Outstanding homes score below the sector average.

The point for an operator is uncomfortable but clarifying. A family asking ChatGPT or Google's AI to recommend a care home is not shown the best-rated homes. The badge those families trust, the one you may have worked years to earn, is invisible to the tool they now ask. AI visibility is a separate discipline from care quality, and right now no one is winning it.

Source: SearchScore SAVI UK Care Homes 2026. 2,196 homes with a current CQC rating at time of audit. Mean GEO Score and share effectively invisible (below 40) by rating band.

The table is set for families, not machines

Why is a well-run home still invisible? Because the signals an assistant weighs before it will name a home are exactly the ones left off the page. A family choosing where a relative will live decides on trust and credentials: a named registered manager, a current rating, genuine reviews, real staff and clear information on fees and services. An AI weighs the same things, but only when they are labelled in a form it can read.

They rarely are. E-E-A-T Content, the evidence of real expertise and experience behind a site, averages just 42.3. Brand Authority averages 33.3, and Platform Optimisation 41.0. Most homes never name the registered manager in machine-readable form, never publish fees, services or a genuine FAQ as structured data, and carry few of the third-party mentions an assistant treats as proof that a real, accountable provider stands behind the home.

None of this is cosmetic. The content usually already exists, written for human readers on an About page or a fees section. It has simply never been relabelled so a machine can confirm it and quote it. That relabelling is a one-off technical job, not a rewrite, and it is where nearly every home in the country is leaving the recommendation on the table.

How care homes compare with other sectors

Care homes are not an outlier that happens to be having a bad year. Across every sector we have benchmarked, the ceiling stays low, and care homes sit at the very bottom of the pack.

Care homes at about 48 are the least AI-visible sector we have measured, roughly level with law firms, and behind dentists and accountants at around 52 and aesthetic clinics at around 56. No sector is doing well. The strategic reading is that in a field where the leader has not yet crossed the line, the bar to become the first genuinely AI-visible care home in your town, or your county, is astonishingly low. Around 33 homes already sit within reach, scoring between 70 and 80, and for them AI-Ready is a weekend of labelling away.

What a care home should actually do

The concrete steps follow straight from the data, in order of leverage:

• Name your registered manager in schema. Add the real, accountable person who runs the home in machine-readable markup, not just on a human-facing team page. It is the honest trust signal a care website should be sending anyway, and it is one an assistant actively looks for.
• Mark the home up as a LocalBusiness or Organisation. Publish the address, the services you provide and the fees as structured data, so an assistant can match the home to a real question like "residential dementia care in Leeds" and quote the facts back.
• Publish the questions families actually ask as FAQ content, about visiting, funding, waiting lists, types of care, and label them in the format AI quotes from. This is the fastest way to lift structured data off the floor from 27.4.
• Add an llms.txt file and let the AI crawlers in, so the assistants building these recommendations can reach and read the site in the first place.
• Earn genuine mentions and reviews on sources an assistant already trusts, directories, local press and the CQC listing, to lift brand authority off 33.3.
• Describe the home consistently everywhere it appears online, so the machine reads one coherent, trustworthy provider.

None of this is a growth hack or a new file format. It is the fundamentals, and this audit is the evidence that in UK care homes they are still completely unclaimed. The gap between the leaders and the rest is a few labels, not the quality of the care.

Frequently asked questions