hcahealthcare.co.uk — AI Site Grade

Analysis

HCA Healthcare UK: AI-Visibility Audit

The site grants every major AI crawler unrestricted 200-level access with full HTML content, yet the cold LLM knowledge about the brand is stale, geographically incomplete, and includes negative billing-practice claims that the site itself never addresses.

Crawler Access

All eleven tested AI bots — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, anthropic-ai, Applebot-Extended, Bytespider, Perplexity-User — receive a 200 status with byte-identical content (~424KB) matching the browser baseline. No UA-based blocking, no Cloudflare challenge, no JS-gating. The site runs on Vercel (x-vercel-id header) with a robots.txt that uses a single User-agent: * rule allowing / and blocking only internal paths (*/data/*, */page-data/*, /settings/*, /undefined/*). No AI-specific directives exist. /llms.txt returns a 404. The sitemap indexes 2,179 URLs plus 2,669 consultant profile URLs in a separate sitemap — strong crawl surface.

Cold-Knowledge Gap

The LLM prior describes HCA UK as operating in "London and Manchester" with hospitals including "The Harley Street Clinic, London Bridge Hospital, and The Wellington Hospital." The site actually lists Birmingham as a third major city (The Harborne Hospital opened 2024) and operates over 30 facilities. The prior mentions "The Portland Hospital for women and children" as an acquisition — the site lists it as a current location but does not frame it as an acquisition. The prior also references "scrutiny over billing practices and aggressive debt collection" — the site contains zero content addressing billing complaints, debt collection, or patient financial disputes. This negative signal in the model's training data has no counter-narrative on the domain.

Schema Posture

The homepage and all top-level pages carry only two schema types: Organization and MedicalOrganization — both with minimal properties (name, url, type). No LocalBusiness, Hospital, MedicalSpecialty, FAQPage, Article, or HealthTopic structured data on the pages that would benefit most. The consultant profile pages are the exception: they use Physician schema with MedicalSpecialty, PostalAddress, EducationalOccupationalCredential, HospitalAffiliation, and BreadcrumbList — this is the site's strongest structured data. The /services page has a BreadcrumbList and a 13-item FAQ section rendered in plain HTML headings (H3s) with no FAQPage schema wrapping them. Blog posts carry only BreadcrumbList — no Article or BlogPosting schema.

Content & Answer Signals

The homepage is text-rich (2,338 words) with clear H1/H2 hierarchy and a strong value proposition ("98% of patients recommend HCA UK", "higher proportion of CQC Outstanding ratings than any other private hospital network"). The blog contains dated articles into 2026 (e.g., April 2026) — content is fresh and clinically detailed. The newsroom shows regular press releases about robotic surgery firsts, clinical trials, and leadership changes. However, the site lacks comparison tables, pricing transparency pages, or structured self-pay cost information that AI engines surface for "how much does private healthcare cost" queries. The /self-pay-paying-for-medical-treatment page exists in the sitemap but was not linked from the homepage navigation.

External Signals

The site links to Doctify for consultant reviews (the sample profile shows 1,005 reviews with a 4.99 rating) and to the CQC provider page. DNS records show integrations with Proofpoint (email security), Sitecore (CMS hints via x-sc-rewrite header), and multiple verification tokens (Apple, Facebook, Google, Cisco, DocuSign). No Reddit threads or major press controversies surfaced in search — the negative billing narrative in the LLM prior appears to come from older training data, not current web presence.

Findings

1. LLM knowledge omits Birmingham and over 30 facilities High

   The LLM prior only lists London and Manchester, missing Birmingham (The Harborne Hospital opened 2024) and over 30 total facilities. This limits AI-generated answers about the full network.

   What to change: Add a structured locations page with schema markup for each hospital and clinic, and ensure the homepage or about page lists all major cities served.

2. Negative billing claims in LLM prior have no counter-narrative on site High

   The LLM prior references scrutiny over billing practices and aggressive debt collection, but the site contains zero content addressing billing complaints, debt collection, or patient financial disputes. This leaves negative signals unchallenged.

   What to change: Publish a billing transparency page or FAQ that addresses common financial concerns, and include schema markup for FAQPage or QAPage.

3. FAQ sections on services page lack FAQPage schema Medium

   The /services page has a 13-item FAQ section rendered as plain HTML headings (H3s) with no FAQPage structured data. This prevents AI from directly surfacing those answers in featured snippets.

   What to change: Wrap each FAQ item in FAQPage schema with Question and Answer properties.

4. Blog posts lack Article or BlogPosting schema Medium

   Blog posts carry only BreadcrumbList schema, missing Article or BlogPosting markup. This reduces the chance of appearing in Google's rich results and AI training signals.

   What to change: Add Article or BlogPosting schema to all blog posts with headline, datePublished, author, and image.

5. No /llms.txt file published Medium

   The site returns a 404 for /llms.txt, missing an opportunity to provide AI crawlers with a curated list of important pages and context.

   What to change: Create an /llms.txt file listing key pages (locations, services, blog, consultant profiles) with brief descriptions.

6. Homepage and top-level pages have minimal schema Medium

   The homepage and all top-level pages carry only Organization and MedicalOrganization schema with minimal properties. No LocalBusiness, Hospital, MedicalSpecialty, or HealthTopic schema is used, limiting AI understanding of the site's scope.

   What to change: Add Hospital, LocalBusiness, and MedicalSpecialty schema to relevant pages, and include address, telephone, and opening hours.

7. No structured self-pay cost information on site Medium

   The site lacks comparison tables, pricing transparency pages, or structured self-pay cost information that AI engines surface for 'how much does private healthcare cost' queries. The /self-pay-paying-for-medical-treatment page exists in the sitemap but is not linked from the homepage navigation.

   What to change: Prominently link the self-pay page from the homepage and add structured data (e.g., Product or PriceSpecification) for common procedures.

8. Robots.txt has no AI-specific directives Low

   The robots.txt uses a single User-agent: * rule and does not name any AI bots. While this allows access, it misses the chance to guide AI crawlers to the most important pages.

   What to change: Add specific directives for GPTBot, ClaudeBot, and other AI crawlers to prioritize key content.

What's working

• All major AI crawlers receive full HTML content with 200 status — Eleven tested AI bots get unrestricted 200-level access with byte-identical HTML content, no JS-gating, and no Cloudflare challenges. This ensures AI crawlers can index the full site.
• Consultant profile pages use comprehensive Physician schema — Consultant profiles include Physician schema with MedicalSpecialty, PostalAddress, EducationalOccupationalCredential, HospitalAffiliation, and BreadcrumbList. This is the site's strongest structured data and helps AI understand specialist details.
• Sitemap indexes over 4,800 URLs including consultant profiles — The sitemap contains 2,179 URLs plus a separate consultant finder sitemap with 2,669 URLs, providing strong crawl surface for AI bots.
• Blog contains fresh, clinically detailed articles into 2026 — The blog has dated articles into April 2026 with detailed clinical content, signaling active content creation that AI can surface for health queries.
• Homepage is text-rich with clear hierarchy and strong value proposition — The homepage has 2,338 words with clear H1/H2 hierarchy and compelling statistics (98% patient recommendation, high CQC ratings), providing strong signals for AI summarization.
• Site links to Doctify and CQC for third-party validation — Consultant profiles link to Doctify (1,005 reviews, 4.99 rating) and CQC provider page, providing external credibility signals that AI can reference.