The SetpointHQ Index · Sector report 3 of 14
The AI assistants are more willing to recommend UK care employers than they are the banks, the retailers or the tech firms. On this week’s scoring run, most of the sector still isn’t converting that goodwill, and the reason is sitting in its own markup.
The SetpointHQ Index scores 100 UK employer brands on how visible they are to AI assistants. We put real candidate questions to ChatGPT, Claude, Gemini and Perplexity and score 5 dimensions: citation (do the assistants name you), schema (does your site give machines structured data to read), knowledge graph (do the machines know who you are), crawlability (can AI crawlers get in) and social footprint. Each week we read one sector closely. Week 1 covered banking, week 2 retail and consumer goods. This week: social care, the five biggest care operators we track.
The full table
| Brand | Composite | Cited | Schema | Knowledge graph | Crawlability | Social |
|---|---|---|---|---|---|---|
| Care UK | 65.7 | 50% | 80 | 59.7 | 75 | 86 |
| HC-One | 57.2 | 40% | 70 | 60.4 | 85 | 64 |
| Barchester Healthcare | 50.5 | 45% | 30 | 69.7 | 75 | 70 |
| Agincare | 45.6 | 40% | 60 | 0 | 68 | 74 |
| Cera Care | 45.6 | 30% | 25 | 59.6 | 75 | 72 |
Composite and dimension columns scored out of 100. Cited is the share of relevant AI answers that name the brand. Care UK ranks 11th of the full 100, Cera Care 75th. Scored in the Index run of 12 July 2026, published 13 July 2026.
Cited more readily than the banks
Start with the good news, because in this sector there is some. Care employers get named by AI more often than employers in categories with far deeper pockets. The median care brand is cited in 40% of relevant answers. Banking sits at 25%, retail at 27.5%, consumer tech at 20%. Only professional services and legal score higher than care. Care UK is named in half of the answers we tested.
That is a real advantage and almost nobody is using it. The assistants are receptive to care as a place to work, no single brand owns that space, and the sector that is shortest of candidates is being handed a way to reach them. Most of it is walking past.
The assistants are ready to recommend care employers. Most of the sector hasn’t made itself easy to recommend.
The schema split
What holds the composite below the Index median is not fame and it is not budget. It is the machine-readable layer, the schema that tells an assistant who you are and what roles you have open, and the sector splits clean down the middle. Care UK and HC-One have built for machines, at 80 and 70. Barchester and Cera Care have barely started, at 30 and 25. That one dimension is the biggest swing in the table.
Cera Care is the sharp end of it, and the detail matters. Cera marks up every job it posts. Open any live Cera vacancy and the JobPosting schema is there, and that job-level markup is the entire reason it scores 25 rather than 0. What Cera never publishes is anything about the employer itself. No Organization schema, no site-level structured data, nothing that describes who Cera is as a place to work. So a machine can read a specific Cera vacancy in full and still can’t describe the company. Ask the four assistants about Cera by name and every one names it. Ask which care employer a candidate should pick and Cera surfaces in 2 of 16 answers. The jobs are legible. The employer isn’t.
Agincare shows the other version of the same hole. Its markup is fine, but it has no knowledge-graph entity at all, no Wikidata record, so it scores a flat 0 there. When a model looks for a neutral account of who Agincare is, there is nothing to find. Different gap, same result: less for the machine to go on, so less it will say.
The cost of closing either gap is small. Organization schema, a claimed knowledge-graph entity, consistent markup across the site. That is a few days of technical work, not a media campaign. Expensive things get budgets and effective things get forgotten.
What this means for a TA leader
First, nobody in UK care owns the AI answer yet, and the assistants are already on your side. The best citation score in the sector still has room above it, the space at the top of the answer is open, and the first mover takes it cheaply.
Second, the gap between the top of this table and the bottom is technical, not financial. Being the largest name in the sector will not carry you if the machines can’t read you, and a smaller operator that has done the markup will pass you in the answer.
Third, marking up your jobs is not the finish line. Job markup gets your vacancies read. Describing the employer, in schema and in the knowledge graph, is what gets you recommended.
On Friday we publish the deep dive on one brand from this table. The finding picks which, not the alphabet.
Where does your brand land?
We will run your employer brand against the 100 on the Index: your 5 dimension scores and the single finding costing you the most applications.
Source: The SetpointHQ Index, 100 UK employer brands. Scored in the run of 12 July 2026. Explore the full Index at pro.setpointhq.com/the-index.