AI Search Audit
TL;DR
The site is in a healthier starting position than most local clinics — it already ships llms.txt, llms-full.txt, MedicalClinic + MedicalBusiness JSON-LD, FAQ schema on the homepage, and an AI-crawler-friendly robots.txt. That covers roughly the top quartile of table-stakes GEO hygiene.
The two biggest gaps vs. April-2026 best practice are authorship / E-E-A-T signals on team and service pages (no Physician schema, no 'medically reviewed by' block, no sameAs links to professional registers), and content coverage — there are no condition-specific landing pages even though the FAQs and service blurbs reference those topics.
Google AI Overviews answer ~89% of healthcare-related queries in 2026, and the 2026 core update explicitly rewards pages with clear, qualified authorship — these are the two changes most likely to move citations.
Already implemented
- llms.txt and llms-full.txt auto-generated from site data
- Explicit AI-crawler allow-list in robots.txt (GPTBot, ClaudeBot, PerplexityBot, Googlebot-Extended, CCBot, anthropic-ai)
- MedicalClinic + MedicalBusiness JSON-LD in base template with address, hours, priceRange, offer catalog
- FAQPage JSON-LD on the homepage
- Canonical URLs, OpenGraph, Twitter Card tags on every page
- Auto-generated sitemap with lastmod
Gaps (reference)
Background context. The action buttons live on the Punch-list below — each punch-list item addresses one or more of these gaps.
G1. No Physician / Person schema on team profiles
YMYL + 2026 core update rewards qualified, machine-readable authorship.
G2. No AggregateRating / Review schema
'4.9 on Google' is on the homepage but not marked up, so AI cannot quote it.
G3. No per-service Service / MedicalTherapy schema
Service pages inherit the site-wide block but emit no page-level schema.
G4. No BreadcrumbList
AI engines use breadcrumbs to understand hierarchy; also unlocks SERP breadcrumbs.
G5. FAQ schema only on homepage
Service-specific Q&As are the most citation-friendly structure in 2026.
G6. No visible dateModified / last-updated
65% of AI citations go to content published in the last year; freshness matters.
G7. No condition-specific pages
No indexable URL for 'sciatica osteopath Christchurch' etc. — biggest content gap.
G8. No blog / editorial content
Freshness cliff at ~3 months; ongoing publishing keeps citable surface area alive.
G9. Homepage H1 not keyword-aligned
'Our Practitioners at Meridian Osteopathy' should be H2; H1 should describe primary intent.
G10. llms.txt could be richer
Missing Conditions-treated / Pricing-ACC / Common-questions sections.
G11. No sameAs / social profile links
Empty social URLs in site.js; nothing for AI engines to disambiguate the entity with.
G12. No MedicalWebPage / ProfilePage per team profile
Each practitioner URL should be a standalone retrievable schema node.
G13. No llms.txt reference in HTML head
Emerging convention; several indexers check for it.
G14. Image alt text could be richer
AI-image-grounded models benefit from descriptive, not titular, alt text.
G15. No off-site citation strategy
SE Ranking (2026): Reddit/Quora brand mentions give ~4× AI citation rate.