dacbeachcroft.com — AI Site Grade

Analysis

DAC Beachcroft — AI-Visibility Audit

The site has zero structured data across every page examined — no Organization, LegalService, Article, Person, or BreadcrumbList schema — despite being a 3,000-person international law firm with 4,700+ URLs in its sitemap. This is the single most consequential gap: AI crawlers can read the content but cannot reliably identify what the entity is, who its people are, or how pages relate.

Crawler Access

All major AI bots — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Bytespider, Applebot-Extended, anthropic-ai — receive a full 200 response with identical byte content to a browser visit. The robots.txt is a single User-agent: * Allow: / rule with no AI-specific directives. The site runs on Microsoft-IIS/10.0 (Sitecore CMS) behind standard security headers including HSTS and CSP. No CDN or WAF layer is detectable. No bot is blocked, throttled, or redirected. The llms.txt returns a 404 — the Sitecore error page renders, not a missing-file redirect.

Cold-Knowledge Gap

The LLM prior knows DAC Beachcroft as a "UK-based international law firm specializing in insurance, health, and litigation" formed from a 2012 merger, with a "Claims Solutions" platform. The actual site positions itself far more broadly: six sectors (Financial Services, Health and Social Care, Insurance, Real Estate, Technology, plus Industry Specialisms), a dedicated AI practice, proprietary tools like AiDan (AI-powered claims assessment), ACE (automated witness statements), and 8DB (in-house barrister service). The cold model knows nothing about AiDan, the Dubai expansion (announced May 2026), the Innovations Lab, or the firm's AI advisory work with the European Parliament. The gap between the model's stale 2012-merger framing and the site's current innovation narrative is wide.

Schema Posture

Zero JSON-LD was found on any page: homepage, sector pages (Insurance, AI, Technology), news articles, people directory, or the "What we think" hub. No Article schema on news items (despite publish dates and author bylines visible in the text). No Person schema on profile pages. No BreadcrumbList despite deep navigation hierarchies. No FAQPage schema despite question-adjacent content patterns. The meta description on the News page and "What we think" page is a generic careers pitch ("Join DAC Beachcroft and build your career...") rather than describing the actual page content — a crawl-quality issue that affects how AI engines classify those pages.

Content & Navigation

The "What we think" hub and the "News" listing page are JavaScript-filter shells — they return ~400 words of visible text consisting almost entirely of navigation markup and a filter bar, with no article previews loaded. Individual article pages (e.g., the Will Sefton hire, Dubai expansion) render full content server-side and are well-written, but they lack any schema markup. The "Our people" directory is similarly a JS-dependent search interface with no people data in the initial HTML. The sitemap contains 4,729 URLs, including a /NotFound page indexed — a crawl-quality issue.

External Signals

The site references Chambers & Partners UK 2026 rankings and Legal 500 recognition in its AI practice page text, but no external review aggregators or press coverage surfaced in search. The firm's external link profile points to esg.dacbeachcroft.com, jobs.dacbeachcroft.com, LinkedIn, and Instagram. The DNS TXT records confirm Sitecore and Mimecast — a standard enterprise stack with no AI-specific optimizations.

Findings

1. Zero JSON-LD structured data across all pages High

   No Organization, LegalService, Article, Person, or BreadcrumbList schema found on any page examined, including the homepage, sector pages, news articles, people directory, and content hub. This prevents AI crawlers from reliably identifying the entity, its people, or page relationships.

   What to change: Add JSON-LD structured data for Organization on the homepage, LegalService on practice pages, Article on news items, Person on profile pages, and BreadcrumbList on all pages.

2. Cold-knowledge gap on innovation narrative and proprietary tools High

   LLM prior knowledge is limited to a 2012 merger and insurance specialization, missing the firm's current six-sector positioning, AI practice, proprietary tools (AiDan, ACE, 8DB), Dubai expansion, and Innovations Lab.

   What to change: Publish an llms.txt file and ensure key innovation pages are richly structured with schema to improve AI knowledge extraction.

3. News and 'What we think' hubs are JavaScript-filter shells High

   The News listing page and 'What we think' hub return only ~400 words of navigation and filter markup, with no article previews in the initial HTML. AI crawlers that do not execute JavaScript see empty shells.

   What to change: Server-render article previews in the initial HTML for listing pages, or provide a static fallback for non-JS crawlers.

4. People directory is a JavaScript-dependent search interface High

   The 'Our people' page returns only 452 words of navigation and a search interface, with no people data in the initial HTML. AI crawlers cannot index individual profiles.

   What to change: Server-render a list of people with links to individual profile pages, or provide a static HTML fallback.

5. llms.txt returns 404 error page Medium

   The llms.txt file is missing, returning a Sitecore error page instead of a redirect or 404. This is a missed opportunity to guide AI crawlers to key content.

   What to change: Create an llms.txt file listing key pages (e.g., AI practice, sectors, news) to improve AI knowledge extraction.

6. Generic meta descriptions on key content pages Medium

   The News page and 'What we think' hub use a generic careers pitch meta description ('Join DAC Beachcroft and build your career...') instead of describing the actual page content, harming AI classification.

   What to change: Update meta descriptions to accurately reflect the page content, e.g., 'Latest news and insights from DAC Beachcroft.'

7. No Article schema on news articles despite visible metadata Medium

   News articles have publish dates and author bylines in the text but lack Article schema, reducing their visibility in AI-powered search results.

   What to change: Add Article schema with headline, datePublished, author, and description to all news articles.

8. No BreadcrumbList schema on deep navigation pages Medium

   Despite deep navigation hierarchies (e.g., /en/What-we-do/Sectors/Technology/Artificial-Intelligence), no BreadcrumbList schema is present, reducing AI understanding of site structure.

   What to change: Add BreadcrumbList schema to all pages with breadcrumb navigation.

9. Sitemap includes /NotFound page Low

   The sitemap contains a /NotFound URL, indicating a crawl-quality issue that wastes crawler budget.

   What to change: Remove the /NotFound URL from the sitemap.

10. No FAQPage schema on question-adjacent content Low

    Pages with question-adjacent content patterns lack FAQPage schema, missing an opportunity for rich results.

    What to change: Identify pages with FAQ-like content and add FAQPage schema.

What's working

• All major AI bots allowed and served full content — All 11 tested AI bots (GPTBot, ClaudeBot, PerplexityBot, etc.) receive a 200 response with identical content to a browser visit. No bot is blocked, throttled, or redirected.
• Individual article pages are server-rendered with full content — News article pages (e.g., Dubai expansion, Will Sefton hire) render full content server-side with well-written text, making them accessible to AI crawlers.
• Sitemap with 4,729 URLs provides broad crawl coverage — The sitemap contains 4,729 URLs, giving crawlers a comprehensive map of the site's content.
• Dedicated AI practice page with detailed content — The AI practice page (1,484 words) covers the firm's AI advisory work, proprietary tools (AiDan, ACE, 8DB), and sector expertise, providing rich content for AI crawlers.
• Standard security headers (HSTS, CSP) in place — The site uses HSTS and CSP headers, indicating a secure configuration that does not hinder AI crawlers.