rula.com — AI Site Grade

Analysis

Rula.com AI-Visibility Audit

The robots.txt explicitly blocks Google-Extended and anthropic-ai from /therapists/ and /providers/ — the site's core directory pages — yet every AI crawler tested (including those two) receives a 200 with full content from those blocked paths, meaning the block is honor-system only and the server does not enforce it.

Crawler Access

All 11 tested AI crawler UAs (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, anthropic-ai, OAI-SearchBot, ChatGPT-User, Bytespider, Applebot-Extended, Perplexity-User) plus a browser baseline return identical 200 status and ~700KB body on the homepage. The /therapists/ny/new-york directory page also serves 200 to every bot with ~1.4MB of content. The site runs on AWS CloudFront with Next.js server-side rendering — no JS-rendering risk, no UA-based firewall. The robots.txt disallow rules for AI bots are unenforced at the server level, meaning compliant crawlers (Google-Extended, anthropic-ai) will voluntarily skip the directory pages while non-compliant ones (GPTBot, ClaudeBot, PerplexityBot) will index them fully. No llms.txt exists (404).

Cold-Knowledge Gap

The LLM knows Rula as a telehealth platform founded in 2020 by "James R." that raised over $100M including an $80M Series B in 2022. The actual site says nothing about founding year, founder name, or funding. The "James R." detail appears to be hallucinated or conflated with another company. The site describes itself as serving 170M+ insured individuals with 20,000+ providers and a 97% match rate — none of these specific metrics appear in the model's cold knowledge. The cold knowledge also mentions "inconsistent provider availability and billing issues" as reputational signals from 2023-2024, but the site prominently features peer-reviewed clinical studies (published in INQUIRY and JMIR Pediatrics, 2025-2026) and a 2025 State of Mental Health Report — evidence of clinical credibility the model is entirely unaware of.

Schema Posture

Every page carries a consistent WebSite + WebPage + BreadcrumbList + ImageObject schema graph. However, the schema is generic: no MedicalBusiness, HealthInsurancePlan, Physician, MedicalWebPage, or FAQPage types are used anywhere. The FAQ page (/faq/) has visible Q&A content but is marked up as a plain WebPage rather than FAQPage with mainEntity — a missed opportunity for AI answer engines to extract structured answers. The WebSite schema includes a SearchAction target, but the template URL (/?s={search_term_string}) does not match the actual site search behavior.

Content & Signals

The homepage is content-rich (2,133 words) with clear H1/H2 structure, an FAQ section, insurance partner logos, and city/state directory links. The blog has regularly published articles with author bylines and dates. The clinical-approach page links to published peer-reviewed studies — a strong trust signal. However, the canonical URL on /clinical-approach/ points to /clincial-approach/ (a typo), and that misspelled URL returns a 404. The sitemap index references three sub-sitemaps covering 1,753+ URLs, including insurance-specific landing pages, therapy-type pages, and city directories — a healthy index footprint.

External Signals

The site links to Trustpilot reviews and social channels (Instagram, LinkedIn, YouTube), but Trustpilot blocks automated access (403). No Reddit threads or press articles surfaced in search results. The DNS TXT records reveal integrations with OpenAI (openai-domain-verification), HubSpot, Calendly, Stripe, and Zoom — indicating active API-level connections to AI and SaaS platforms. The OpenAI verification record suggests Rula has already engaged with OpenAI's platform, yet the site has no llms.txt to guide that integration.

Findings

1. Robots.txt blocks AI crawlers from directory pages but server does not enforce High

   The robots.txt disallows Google-Extended and anthropic-ai from /therapists/ and /providers/, yet all tested AI crawlers receive a 200 with full content from those paths. Compliant bots will voluntarily skip the pages while non-compliant ones index them fully.

   What to change: Either remove the disallow rules from robots.txt to allow all AI crawlers, or implement server-side access control (e.g., IP-based blocking or token authentication) to enforce the block.

2. No llms.txt file found Medium

   The site returns a 404 for /llms.txt, missing an opportunity to guide AI crawlers to key resources. An OpenAI domain verification TXT record exists, indicating active engagement with AI platforms.

   What to change: Create an llms.txt file listing key pages (homepage, about, clinical approach, FAQ, blog) and the sitemap URL.

3. Schema markup uses generic types, missing healthcare-specific types Medium

   All pages use WebSite, WebPage, BreadcrumbList, and ImageObject but lack MedicalBusiness, Physician, HealthInsurancePlan, MedicalWebPage, or FAQPage types. The FAQ page has visible Q&A content but is marked as WebPage instead of FAQPage with mainEntity.

   What to change: Add FAQPage schema to the FAQ page with mainEntity for each Q&A. Add MedicalBusiness and Physician schema to provider and directory pages. Consider adding HealthInsurancePlan schema to insurance landing pages.

4. Canonical URL points to a misspelled 404 page High

   The clinical-approach page's canonical URL is /clincial-approach/ (typo), which returns a 404. This confuses search engines and undermines the page's authority.

   What to change: Update the canonical URL on /clinical-approach/ to point to itself (https://www.rula.com/clinical-approach/). Set up a 301 redirect from /clincial-approach/ to /clinical-approach/.

5. LLM cold knowledge lacks key site metrics and clinical credibility signals Medium

   The LLM's prior knowledge of Rula is sparse and partially hallucinated (founder name, funding details). The site's specific claims (170M+ insured, 20,000+ providers, 97% match rate, peer-reviewed studies) are absent from the model's knowledge.

   What to change: Ensure key metrics and clinical credibility signals are prominently featured in structured data and on high-authority pages. Consider publishing an llms.txt to guide AI crawlers to these facts.

6. WebSite SearchAction template URL does not match actual site search Low

   The schema's SearchAction target uses /?s={search_term_string}, but the site's search behavior may differ. This can mislead AI agents trying to use site search.

   What to change: Verify the actual search URL pattern and update the SearchAction target accordingly.

7. Trustpilot reviews page blocks automated access Low

   The Trustpilot review page returns a 403 for automated requests, preventing AI crawlers from accessing third-party review content that could enhance credibility.

   What to change: No fix available on Rula's side; consider embedding Trustpilot widgets or summaries on the site to make review content crawlable.

What's working

• All AI crawlers receive 200 status on homepage and directory pages — Every tested AI crawler gets a full 200 response with server-rendered content on the homepage and directory pages, ensuring content is accessible to compliant bots.
• Homepage is content-rich with clear structure and FAQ section — The homepage contains over 2,000 words with H1/H2 hierarchy, an FAQ section, insurance partner logos, and city/state directory links, providing strong topical signals.
• Blog publishes regularly with author bylines and dates — The blog has frequently updated articles with author attribution and publication dates, signaling fresh, authoritative content to search engines and AI crawlers.
• Clinical approach page links to published peer-reviewed studies — The clinical-approach page references studies in INQUIRY and JMIR Pediatrics (2025-2026), providing strong trust signals for AI models evaluating credibility.
• Sitemap index covers over 1,750 URLs across multiple sub-sitemaps — The sitemap index references three sub-sitemaps with 1,753+ URLs, including insurance, therapy-type, and city directory pages, ensuring broad crawl coverage.
• Every page carries consistent WebSite, WebPage, BreadcrumbList, and ImageObject schema — The site has a uniform schema foundation across all pages, providing basic structured data for search engines and AI crawlers.
• DNS TXT record confirms OpenAI domain verification — The presence of an openai-domain-verification TXT record indicates Rula has an active integration with OpenAI, suggesting readiness for AI-driven features.
• 2025 State of Mental Health Report published on site — The site hosts a dedicated page for its 2025 State of Mental Health Report, a data-rich asset that can serve as a citation source for AI models.