compexlegal.com — AI Site Grade

Analysis

Compex Legal Services — AI-Visibility Audit

The site publishes blog posts dated May 2026 (future-dated content) that AI crawlers can freely index, yet the homepage and all key pages carry zero JSON-LD schema — no Organization, no LocalBusiness, no Product — despite the site being a 50-year-old company with 17+ offices and an AI product called Asabell.

Crawler Access

All major AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Bytespider, Applebot-Extended — receive a full 200 response with identical byte payload (323,785 bytes) to a browser baseline. No UA-based blocking exists. The robots.txt is minimal (5 HubSpot admin-path disallows, no AI-bot rules) and llms.txt returns a 404 — the site has no AI-friendly content map. The site runs on HubSpot behind Cloudflare, with an openai-domain-verification TXT record confirming OpenAI has already verified the domain for potential GPT integration, yet no structured guidance for AI crawlers exists.

Cold-Knowledge Gap

The LLM knows Compex as a "30+ year" medical record retrieval company focused on workers' comp and personal injury, with a California-centric reputation and "occasional delays" mentioned in reviews. The actual site claims "founded in 1972" (53+ years), positions itself as "the #1 Medical Record Retrieval Service" and "the largest national procurer of medical records" with "over 5,000 Law Firms & 350 Insurance Carriers" as clients, and heavily promotes Asabell — an AI-powered summarization tool that the model's cold knowledge does not mention at all. The model also references "California" as a primary geography, while the site lists offices in 17+ locations across 9 states and claims nationwide 50-state service.

Schema Posture

The homepage, about page, solutions pages, and FAQ page all contain zero JSON-LD schema blocks. Only individual blog posts carry BlogPosting schema with Organization publisher data. The FAQ page — which has 18+ well-structured Q&A pairs — uses plain HTML headings with no FAQPage schema. The about page lists 8 C-suite executives with names and titles but no Person schema. The technology page describes a product (Asabell) with specific claims (98% accuracy, 67% less review time) but no Product or SoftwareApplication schema.

External Signals

External search results for Compex are near-zero — no Reddit threads, no review aggregator results, no press coverage surfaced in standard web search. The site's external link profile is limited to LinkedIn, YouTube, and an ADP careers portal. The blog is the primary content engine, with posts dating from February 2026 through May 28, 2026 (a date that has not yet occurred at time of audit), suggesting either aggressive forward-scheduling or a content-date configuration issue that could confuse temporal signals for AI engines.

Findings

1. Zero JSON-LD on all core pages — Homepage, about, solutions, FAQ, technology pages all lack Organization, FAQPage, Product, LocalBusiness, or Person schema. Blog posts are the only pages with any structured data (BlogPosting). 2. No llms.txt — Returns 404. No AI-friendly content map exists despite OpenAI domain verification being active. 3. robots.txt has no AI-bot directives — No GPTBot, ClaudeBot, or PerplexityBot rules. All AI crawlers are implicitly allowed but receive no guidance on crawl frequency or preferred paths. 4. Future-dated blog content — Multiple posts carry datePublished values in May 2026 (e.g., 2026-05-28), which is in the future relative to the current date. This may degrade temporal credibility signals for AI models. 5. Cold-knowledge gap on Asabell — The LLM's prior knowledge does not mention Compex's flagship AI product (Asabell), despite the site heavily promoting it as a differentiator. 6. FAQ page has 18+ Q&A pairs with no FAQPage schema — A strong answer-format signal is invisible to AI crawlers looking for structured FAQ markup. 7. About page lists 8 executives with no Person schema — Leadership names and titles are plain text only. 8. Contradictory office count — The about page states "17 offices" in one section and "36 offices" in another, both referencing the same headquarters location. 9. Near-zero external citation footprint — No Reddit, review sites, or press coverage surfaced in web search, limiting the off-domain signals AI engines use to validate brand authority. 10. HubSpot-hosted with Cloudflare — Site uses x-hs-cache-config and HubSpot-specific headers. The 323KB page payload is large for AI crawler processing efficiency.

Findings

1. Zero JSON-LD schema on all core pages High

   Homepage, about, solutions, FAQ, and technology pages lack Organization, FAQPage, Product, LocalBusiness, or Person schema. Only blog posts have BlogPosting schema.

   What to change: Add Organization schema to the homepage, FAQPage schema to the FAQ page, Product or SoftwareApplication schema to the Asabell solution page, and Person schema for executives on the about page.

2. No llms.txt file (404) High

   The site returns a 404 for llms.txt, providing no AI-friendly content map despite OpenAI domain verification being active.

   What to change: Create an llms.txt file listing key pages and summaries to guide AI crawlers.

3. Robots.txt has no AI-bot directives Medium

   The robots.txt file only disallows HubSpot admin paths and does not include rules for GPTBot, ClaudeBot, or PerplexityBot, leaving AI crawlers unguided.

   What to change: Add explicit directives for AI crawlers, such as crawl-delay and allowed paths, to manage crawl behavior.

4. Future-dated blog content Medium

   Multiple blog posts have datePublished values in May 2026 (e.g., 2026-05-28), which is in the future relative to the current date, potentially degrading temporal credibility signals for AI models.

   What to change: Ensure all blog post dates reflect actual publication dates and avoid future-dated content to maintain temporal accuracy.

5. Cold-knowledge gap on Asabell AI product Medium

   The LLM's prior knowledge does not mention Asabell, Compex's flagship AI-powered medical record summarization tool, despite the site heavily promoting it.

   What to change: Increase external visibility of Asabell through press releases, case studies, and structured data to help AI models associate the product with the brand.

6. FAQ page has 18+ Q&A pairs with no FAQPage schema High

   The FAQ page contains well-structured Q&A pairs in plain HTML headings but lacks FAQPage schema, making the content invisible to AI crawlers looking for structured FAQ markup.

   What to change: Add FAQPage schema with mainEntity array to the FAQ page to enable rich results and AI-friendly extraction.

7. About page lists 8 executives with no Person schema Medium

   The about page lists C-suite executives with names and titles but no Person schema, missing an opportunity for AI models to recognize leadership.

   What to change: Add Person schema for each executive with name, jobTitle, and sameAs links.

8. Contradictory office count on about page Low

   The about page states '17 offices' in one section and '36 offices' in another, both referencing the same headquarters location, creating inconsistency.

   What to change: Reconcile the office count to a single accurate number across the site.

9. Near-zero external citation footprint Medium

   Web searches for Compex Legal Services on Reddit, review sites, and press coverage returned no results, limiting off-domain signals AI engines use to validate brand authority.

   What to change: Build external citations through press releases, industry awards, and active participation in legal tech forums.

10. Large page payload (323KB) for AI crawlers Low

    The homepage serves 323,785 bytes, which is large for AI crawler processing efficiency, potentially slowing down indexing.

    What to change: Optimize page size by minifying assets and deferring non-critical resources to reduce payload for crawlers.

What's working

• All major AI crawlers receive full access — GPTBot, ClaudeBot, PerplexityBot, and others receive a 200 response with identical content to browser users, ensuring no blocking.
• OpenAI domain verification TXT record present — The DNS TXT records include an openai-domain-verification record, indicating OpenAI has verified the domain for potential GPT integration.
• Blog posts include BlogPosting schema — Individual blog posts carry BlogPosting schema with Organization publisher data, providing some structured data for AI models.
• Sitemap available with 80 URLs — The sitemap returns 200 with 80 URLs, providing a clear content inventory for crawlers.
• HubSpot hosting with Cloudflare CDN — The site uses HubSpot CMS behind Cloudflare, providing reliable hosting and CDN caching for performance.