edvoyles.com — AI Site Grade

Analysis

Cloudflare 403 Wall: Ed Voyles Automotive Group is Invisible to Every AI Crawler

Every major AI crawler — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Applebot-Extended, Bytespider, and anthropic-ai — receives a 403 Forbidden from www.edvoyles.com, identical to the block served to a standard browser. The site sits behind Cloudflare with no robots.txt, no llms.txt, no accessible sitemap, and no Wayback Machine snapshot. The only visible page title is "Dealer Website" — a generic CMS shell that never renders content to any automated agent.

Crawler Access

All 11 bot user-agents tested returned 403 with 4,568 bytes (Cloudflare challenge page), while the baseline browser fetch returned 236,655 bytes of the same blocked HTML. The robots.txt and llms.txt endpoints also return 403 with the same Cloudflare HTML shell — neither file exists or is served. The domain resolves to a single IP (54.243.57.127) behind Cloudflare, with nameservers at Network Solutions (worldnic.com). No subdomains, no alternate paths, and no .well-known/ resources are accessible.

Cold-Knowledge Gap

The LLM model describes Ed Voyles Automotive Group as a family-owned Marietta, Georgia dealership group selling Honda, Hyundai, Kia, and CDJR brands, founded in the 1960s by Ed Voyles Sr. This knowledge exists entirely from off-domain training data — the model has never seen the actual website. The site itself provides zero verifiable content to confirm or update this prior. Any changes in inventory, brand representation, location, or ownership since the training data cutoff are invisible to AI systems.

External Signals

A DuckDuckGo search for "Ed Voyles", edvoyles.com, Ed Voyles Honda Marietta, and Ed Voyles Automotive Group returned zero results across all queries. The brand has no discoverable external mentions, reviews, Reddit threads, press coverage, or directory listings in the search index queried. This total absence of external citation means AI models have no corroborating signals to validate or enrich the cold knowledge they hold.

Schema Posture

The homepage returns zero JSON-LD schema, zero structured data of any kind. No Organization, AutoDealer, LocalBusiness, or Product schema is present. The page has no meta description, no Open Graph tags, no canonical URL, and no headings. The entire site is a blank slate to any parser that cannot execute JavaScript — and even JS-rendering tools would hit the Cloudflare 403 before reaching any application code.

Findings

1. Cloudflare 403 blocks all AI crawlers High

   Every major AI crawler (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Applebot-Extended, Bytespider, anthropic-ai) receives a 403 Forbidden response from www.edvoyles.com, identical to the block served to a standard browser. The site sits behind Cloudflare with no accessible content for any automated agent.

   What to change: Configure Cloudflare to allow AI crawler user-agents (e.g., GPTBot, ClaudeBot) by updating WAF rules or creating a firewall bypass for known bot IP ranges. Alternatively, serve a static HTML version of the site to bots.

2. No robots.txt file accessible High

   The robots.txt endpoint returns a 403 Forbidden with a Cloudflare HTML shell, meaning no robots.txt file exists or is served. This prevents crawlers from understanding which paths are allowed or disallowed.

   What to change: Create and serve a robots.txt file that allows AI crawlers (e.g., GPTBot, ClaudeBot) to access the site, and disallow any sensitive paths.

3. No llms.txt file accessible Medium

   The llms.txt endpoint returns a 403 Forbidden, indicating no llms.txt file exists. This file would help AI models discover key content and context about the site.

   What to change: Create an llms.txt file that provides a summary of the business, key pages, and structured data hints for AI crawlers.

4. No sitemap.xml accessible High

   The sitemap.xml endpoint returns a 403 Forbidden, preventing search engines and AI crawlers from discovering the site's pages.

   What to change: Generate and serve a sitemap.xml listing all public pages, and ensure it is accessible to crawlers.

5. Zero JSON-LD schema on homepage High

   The homepage contains no JSON-LD structured data, no Organization, AutoDealer, LocalBusiness, or Product schema. This means AI models cannot extract business details, inventory, or location information from the site.

   What to change: Add JSON-LD structured data for Organization, AutoDealer, LocalBusiness, and Product (for vehicles) to the homepage and relevant pages.

6. No meta description, OG tags, or headings Medium

   The homepage has no meta description, Open Graph tags, canonical URL, or headings. This severely limits how search engines and AI models interpret the page's content and purpose.

   What to change: Add a meta description, Open Graph tags, a canonical URL, and proper heading structure (h1, h2) to the homepage.

7. Zero external search results for brand High

   DuckDuckGo searches for 'Ed Voyles', 'edvoyles.com', 'Ed Voyles Honda Marietta', and 'Ed Voyles Automotive Group' returned zero results. The brand has no discoverable external mentions, reviews, or directory listings in the search index queried.

   What to change: Build external citations by claiming business listings on Google Business Profile, Yelp, DealerRater, and other automotive directories. Encourage customer reviews and press coverage.

8. LLM cold knowledge unverifiable from site Medium

   The LLM describes Ed Voyles Automotive Group as a family-owned Marietta dealership selling Honda, Hyundai, Kia, and CDJR brands, but this knowledge comes entirely from off-domain training data. The site provides zero verifiable content to confirm or update this prior, so any changes since training are invisible.

   What to change: Ensure the website contains accurate, up-to-date information about the business (brands, location, history) and make it accessible to crawlers so AI models can verify and refresh their knowledge.

9. Homepage is a generic CMS shell Medium

   The only visible page title is 'Dealer Website' — a generic CMS shell that never renders content to any automated agent. The page returns 236,655 bytes of blocked HTML to a browser but no meaningful content to crawlers.

   What to change: Replace the generic CMS shell with a custom homepage that includes business name, location, inventory highlights, and structured data.

10. No Wayback Machine snapshot available Low

    The Wayback Machine has no snapshot of www.edvoyles.com, indicating the site has never been archived or was blocked from archiving. This erases historical content and prevents AI models from accessing past versions.

    What to change: Allow the Internet Archive to crawl the site by removing any blocking in robots.txt or Cloudflare settings.

11. No .well-known/ resources accessible Low

    The .well-known/ directory returns a 403 Forbidden, preventing access to standard resource discovery files like security.txt or ai-plugin.json.

    What to change: Serve appropriate .well-known/ resources (e.g., security.txt) and ensure they are accessible to crawlers.

12. No subdomains or alternate paths discovered Low

    No subdomains or alternate paths were discovered for the domain, limiting the site's surface area for AI visibility.

    What to change: Consider adding subdomains for different brands or locations (e.g., honda.edvoyles.com) and ensure they are accessible.