freeths.co.uk — AI Site Grade

Analysis

Search seems to be returning no results across the board. That's a limitation of the search tool right now, not a finding about the domain. I have enough data to write the audit. Let me compile.

Here's the audit:

---

Freeths has no structured data markup anywhere on the site — zero JSON-LD schema on the homepage, about page, people directory, individual lawyer profiles, or the Post Office Horizon scandal landing page — despite being a top-50 UK law firm with 762 listed lawyers, 13 offices, and a landmark national legal case.

Crawler Access

All major AI crawlers reach the homepage with a full 200 response — GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, ChatGPT-User, Applebot-Extended, Bytespider, and anthropic-ai all return the same ~521 KB payload as a browser. Google-Extended is the sole exception, receiving a 403 Cloudflare block with a "Attention Required" challenge page. The robots.txt is minimal (129 bytes, one catch-all rule) and mentions no AI bots by name. No llms.txt exists (404). The site runs on Cloudflare with a strict CSP and HSTS preload.

Schema Posture

No page examined — homepage, /about-us/, /legal-services/, /people/, individual /people/mohammed-abbas/, or the Post Office scandal page — contains any JSON-LD schema of any type. A law firm of this size with 762 people profiles, practice area pages, and a high-profile public-interest case has no LegalService, Organization, Person, FAQPage, Article, or BreadcrumbList markup. This is a complete structured-data vacuum. The people directory lists 762 lawyers with names, roles, and locations but each profile page is a plain HTML document with no Person schema, no knowsAbout, no alumniOf, no worksFor.

Cold-Knowledge Gap

The cold LLM knows Freeths as a "mid-tier" UK firm with "strong regional coverage" focused on public sector procurement, PFI/PPP contracts, and equal pay claims. It mentions the "Freeths Foundation" charity. The cold model has no awareness of the Post Office Horizon scandal work — the single most differentiating fact about the firm. The site itself leads with this case prominently (H1 on homepage, dedicated landing page, "landmark High Court victory for 555 sub-postmasters"). The cold model also does not know Freeths is a B Corp, that it won Law Firm of the Year at both the Legal Business Awards 2024 and City AM Awards 2025, or that it has 13 offices and 1,300+ staff. The site claims "largest B Corp law firm in the UK" — this positioning is entirely absent from the model's prior.

Content Signals

The homepage uses strong answer-format signals: a list of office locations, a news feed with dated articles, and comparison language ("big firm, without the attitude"). However, there are no FAQ sections, no comparison tables, and no structured data to reinforce these signals for AI answer engines. The Post Office scandal page is well-written with clear scheme breakdowns (Horizon Family Members Redress Scheme, Horizon Shortfall Scheme Appeals, Horizon Convictions Redress Scheme) but lacks FAQ markup or WebPage schema. The sitemap contains 4,579 URLs — a large content surface — but none of it is surfaced to AI crawlers through an llms.txt or structured sitemap annotations.

Findings

1. No structured data markup on any page examined High

   The homepage, about page, legal services page, people directory, individual lawyer profile, and Post Office Horizon scandal page all lack any JSON-LD schema. This includes missing LegalService, Organization, Person, FAQPage, Article, and BreadcrumbList markup.

   What to change: Add JSON-LD schema for LegalService on practice area pages, Organization on the homepage, Person on lawyer profiles, FAQPage on the Post Office scandal page, and BreadcrumbList site-wide.

2. Google-Extended blocked by Cloudflare challenge page High

   Google-Extended receives a 403 response with a Cloudflare challenge page, while all other major AI crawlers (GPTBot, ClaudeBot, PerplexityBot, etc.) access the homepage normally. This blocks Google's AI training and retrieval from indexing the site.

   What to change: Allow Google-Extended through Cloudflare by adding an allow rule or removing the challenge for that user-agent.

3. Cold LLM lacks awareness of Post Office Horizon scandal work and B Corp status High

   The cold LLM describes Freeths as a mid-tier firm with regional coverage but has no knowledge of the Post Office Horizon scandal case, B Corp certification, Law Firm of the Year awards, or the firm's size (13 offices, 1,300+ staff). This content is present on the site but not surfaced to AI models.

   What to change: Add structured data (e.g., schema.org/WebPage with about and mentions) to the Post Office scandal page and B Corp page, and publish an llms.txt file to guide AI crawlers to key content.

4. No llms.txt file published Medium

   The site returns a 404 for /llms.txt, meaning there is no machine-readable file to guide AI crawlers to important pages or provide a summary of the site's content.

   What to change: Create an llms.txt file listing key pages (e.g., homepage, practice areas, people directory, Post Office scandal page) and a brief summary of the firm.

5. Robots.txt does not reference any AI bots Low

   The robots.txt file is minimal (129 bytes) with a single catch-all rule and does not explicitly allow or disallow any AI crawlers, leaving their access to default behavior.

   What to change: Add explicit rules for AI crawlers (e.g., GPTBot, ClaudeBot) to ensure desired access and prevent accidental blocking.

6. Lawyer profiles lack Person schema Medium

   The people directory lists 762 lawyers, but individual profile pages (e.g., Mohammed Abbas) contain no Person schema, missing opportunities for AI models to extract expertise, role, and location data.

   What to change: Add Person schema to each lawyer profile with properties such as name, jobTitle, worksFor, knowsAbout, alumniOf, and location.

7. Post Office Horizon scandal page lacks FAQ markup Medium

   The page clearly describes three redress schemes (Horizon Family Members Redress Scheme, Horizon Shortfall Scheme Appeals, Horizon Convictions Redress Scheme) in a question-answer format but has no FAQPage schema.

   What to change: Add FAQPage schema with the three redress schemes as questions and their descriptions as answers.

8. Homepage missing Organization schema Medium

   The homepage does not include Organization schema to convey the firm's name, logo, description, number of employees, office locations, or awards.

   What to change: Add Organization schema to the homepage with properties such as name, logo, description, numberOfEmployees, location, and award.

9. No BreadcrumbList schema on any page Low

   No page examined includes BreadcrumbList schema, which helps AI crawlers understand site structure and context.

   What to change: Add BreadcrumbList schema to all pages to indicate site hierarchy.

10. Legal services page lacks LegalService schema Medium

    The legal services listing page describes practice areas but has no LegalService schema to mark up individual services.

    What to change: Add LegalService schema for each practice area with areaServed, description, and provider.

11. Insights and events page lacks Article schema Low

    The insights and events page lists articles and press releases but has no Article schema to mark up individual posts.

    What to change: Add Article schema to each insight or event post with headline, datePublished, author, and description.

12. Large sitemap lacks structured annotations Low

    The sitemap contains 4,579 URLs but has no annotations (e.g., lastmod, changefreq, priority) to help crawlers prioritize content.

    What to change: Add lastmod and priority annotations to the sitemap to guide crawlers to fresh and important content.

What's working

• All major AI crawlers except Google-Extended access the homepage — GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, ChatGPT-User, Applebot-Extended, Bytespider, and anthropic-ai all receive a 200 response with full content, ensuring broad AI visibility.
• Post Office Horizon scandal page is well-written with clear scheme breakdowns — The page clearly describes three redress schemes and uses strong answer-format signals, making it suitable for AI answer engines if schema were added.
• Homepage uses answer-format signals like office locations and news feed — The homepage includes a list of office locations, a news feed with dated articles, and comparison language, which can help AI models extract structured information.
• Sitemap contains 4,579 URLs indicating a large content surface — The large number of URLs suggests extensive content across practice areas, people profiles, and insights, providing a rich source for AI indexing.
• People directory lists 762 lawyers with names, roles, and locations — The directory provides a comprehensive list of lawyers, which could be enhanced with Person schema to become a valuable AI resource.
• Site uses Cloudflare with strict CSP and HSTS preload — The site has strong security headers, which protect against attacks and ensure safe browsing, though they may also affect crawler access.
• Robots.txt allows all crawlers with no disallowed paths — The minimal robots.txt does not block any crawler, leaving the entire site accessible by default.