gratanet.com — AI Site Grade

Analysis

GRATA International runs a Nuxt.js site on Google Cloud (35.195.49.205) with no CDN or WAF in front — all AI crawlers get a full 200 with real content, no blocks, no JS-shell risk. But the site has zero structured data on any page examined, a broken /awards page that is linked from the sitemap and internal navigation, and no llms.txt. The cold-knowledge gap is significant: the LLM thinks GRATA was founded in 1992 with 20+ offices across 15 countries, while the site itself claims founding in the 1990s, 27 countries, 260+ professionals, and a Swiss Verein structure established in 2018.

Crawler Access

All 11 tested AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Bytespider, Applebot-Extended, anthropic-ai, ChatGPT-User, Perplexity-User) receive 200 status with ~542KB of full HTML content — identical to a browser baseline. No UA-based blocking, no Cloudflare challenge, no JS shell. The robots.txt contains a User-Agent: * rule that only disallows /laravel-filemanager/, /web_files/, /storage/, and /_nuxt/ — AI bots are not mentioned anywhere. The llms.txt returns 404. The site runs on Google Cloud (GCP) with no detectable CDN or WAF layer; the X-Powered-By: Nuxt header confirms a server-rendered Vue.js framework, meaning content is available to crawlers without JS execution.

Schema Posture

Zero JSON-LD or microdata was found on any of the 10+ pages examined — homepage, about, team, career, recognition, association, pro-bono, employee profiles, and insights. No Organization, LegalService, Person, Article, FAQPage, or BreadcrumbList schema exists. The site has rich structured data opportunities: 260+ professionals across 27 countries, rankings from Chambers & Partners and Legal 500, a Swiss Verein association model, and detailed employee bios with education and practice areas — none of it marked up. The /awards page returns 404 despite being listed in the sitemap and linked from the about page.

Cold-Knowledge Gap

The LLM's prior knowledge describes GRATA as "founded in 1992 in Kazakhstan" with "20+ offices across 15 countries" and a "GRATA Legal" platform. The actual site says 27 countries, 260+ professionals, 34 years of experience (implying founding ~1992, consistent), and a Swiss Verein association structure founded in Zug in 2018. The model knows nothing about the association model, the "best friends" office category, the AIFC practice, or the Chairman Andrew Gamble's Hogan Lovells background. The model also references a "GRATA Legal" platform that does not appear on the site — this may be a hallucination or an outdated reference.

External Signals

The site links to LinkedIn, YouTube, Telegram, Facebook, Twitter/X, and Instagram, but no external reviews, Reddit threads, or press coverage surfaced in search. The DNS TXT records show five Google Site Verification tokens (suggesting multiple GSC property migrations or team changes) and a SendPulse + Mailganer email stack. The site has 8 language-specific sitemaps (en, ar, mn, ru, kk, zh, tr, hy) with over 11,000 URLs indexed — a large content footprint for a firm with no schema markup.

Findings

1. Zero JSON-LD or microdata on any page High

   No structured data (Organization, LegalService, Person, Article, FAQPage, BreadcrumbList) was found on any of the 10+ pages examined, missing opportunities to enhance AI understanding of the firm's professionals, offices, and rankings.

   What to change: Add JSON-LD structured data for Organization, LegalService, Person (for each employee), Article (for insights), and BreadcrumbList across all pages.

2. Broken /awards page linked from sitemap and navigation High

   The /awards page returns a 404 error despite being listed in the sitemap and linked from the about page, creating a dead end for crawlers and users.

   What to change: Restore the /awards page with content or remove it from the sitemap and internal links.

3. No llms.txt file available Medium

   The llms.txt endpoint returns a 404, missing an opportunity to provide AI crawlers with a structured overview of the site's content.

   What to change: Create an llms.txt file listing key pages and a brief description of the firm.

4. Cold-knowledge gap misrepresents firm size and structure High

   LLMs describe GRATA as having 20+ offices across 15 countries, while the site claims 27 countries and 260+ professionals. The Swiss Verein association model and 'best friends' office category are absent from LLM knowledge.

   What to change: Add structured data and publish an llms.txt to help AI models accurately represent the firm's global presence and structure.

5. No external reviews or press coverage found in search Medium

   Searches for GRATA International on Reddit, Chambers, Legal 500, and general reviews returned zero results, limiting external signals that could boost AI visibility.

   What to change: Encourage clients and partners to leave reviews on platforms like Google and Clutch, and pursue press coverage to build external signals.

6. No CDN or WAF layer detected Low

   The site runs directly on Google Cloud without a CDN or WAF, which may impact performance and security but does not currently block AI crawlers.

   What to change: Consider adding a CDN like Cloudflare to improve load times and add a WAF for security.

What's working

• All AI crawlers receive full HTML content — All 11 tested AI crawlers receive a 200 status with full HTML content, identical to a browser baseline, with no UA-based blocking or JS shell.
• Server-rendered Nuxt.js provides content without JS execution — The site uses Nuxt.js with server-side rendering, confirmed by the X-Powered-By header, ensuring content is available to crawlers without requiring JavaScript execution.
• Robots.txt allows all AI bots without restrictions — The robots.txt file has a User-Agent: * rule that only disallows a few administrative paths, leaving all AI bots unrestricted access to the site.
• Large content footprint with 11,000+ indexed URLs — The site has 8 language-specific sitemaps with over 11,000 URLs indexed, providing a substantial content base for AI crawlers to discover.
• Social media links to multiple platforms — The site links to LinkedIn, YouTube, Telegram, Facebook, Twitter/X, and Instagram, providing channels for external signals.