dealerbuilt.com — AI Site Grade

Analysis

Cloudflare Wall Blocks Every AI Crawler and Human Browser Alike

Every single AI crawler tested — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, ChatGPT-User, OAI-SearchBot, Applebot-Extended, anthropic-ai, and Bytespider — receives a 403 Cloudflare JS challenge on the homepage, robots.txt, sitemap.xml, and llms.txt. No bot or browser (including baseline) can access a single byte of real content. The site is hosted on Cloudflare (A records at 104.18.14.64/15, NS on bella.ns.cloudflare.com/hunts.ns.cloudflare.com) with Cloudflare's "Under Attack" or JS-challenge mode enabled globally.

Cold-Knowledge Gap: LLM Knows More Than the Site Shows

The model's cold knowledge describes DealerBuilt as a cloud-based DMS provider founded in 2004 in Mason City, Iowa, with a flagship product called "LightYear DMS", a CRM module, digital retailing tools, and a "no hidden fees" pricing model. It also cites generally positive reviews on G2 and Capterra. The actual site at dealerbuilt.com reveals none of this — no homepage text, no product pages, no pricing, no about page, no schema, no headings, no FAQ, no tables. The gap between what the LLM knows and what the site exposes is total: the site is a blank wall.

No External Indexing Footprint

DuckDuckGo returns zero search results for the exact brand name "DealerBuilt", for "DealerBuilt DMS", for "dealerbuilt.com", or for "LightYear DMS". No G2 or Capterra review pages appear in search results. No Wayback Machine snapshot exists. The site has no discoverable subdomains, no indexed URLs, and no external mentions surfaced through web search. The DNS TXT records include an openai-domain-verification token and an anthropic-domain-verification token, indicating the company has proactively registered with both OpenAI and Anthropic for crawler access — yet both GPTBot and ClaudeBot are blocked at the Cloudflare layer anyway.

Schema and Content Signals: Absent

The homepage returns zero JSON-LD schemas, zero structured data types, zero headings, and zero visible text. The llms.txt endpoint returns a Cloudflare challenge (403), not a file. The robots.txt is unreachable. The sitemap.xml is unreachable. There is no FAQ, no comparison table, no list, no definition pattern — no answer-format signals of any kind. The site is a JavaScript shell that requires a Cloudflare Turnstile challenge to proceed, and no automated agent (bot or browser) in this investigation passed it.

Findings

1. Cloudflare JS challenge blocks every AI crawler and browser High

   All tested AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, ChatGPT-User, OAI-SearchBot, Applebot-Extended, anthropic-ai, Bytespider) and baseline browsers receive a 403 Cloudflare JS challenge on the homepage, robots.txt, sitemap.xml, and llms.txt. No bot or browser can access any real content.

   What to change: Disable Cloudflare JS challenge for known AI crawler user agents and allow access to robots.txt, sitemap.xml, and llms.txt without challenge.

2. robots.txt returns 403 Cloudflare challenge High

   The robots.txt file at dealerbuilt.com/robots.txt is blocked by Cloudflare, returning a 403 JS challenge. No user-agent directives are visible, and AI crawlers cannot read any crawling rules.

   What to change: Allow robots.txt to be served without Cloudflare challenge so crawlers can read directives.

3. sitemap.xml returns 403 Cloudflare challenge High

   The sitemap.xml at dealerbuilt.com/sitemap.xml is blocked by Cloudflare, returning a 403 JS challenge. Search engines cannot discover site URLs.

   What to change: Allow sitemap.xml to be served without Cloudflare challenge.

4. llms.txt returns 403 Cloudflare challenge High

   The llms.txt endpoint at dealerbuilt.com/llms.txt is blocked by Cloudflare, returning a 403 JS challenge. AI crawlers cannot access the intended LLM guidance file.

   What to change: Allow llms.txt to be served without Cloudflare challenge.

5. Homepage returns no visible text or structured data High

   The homepage at dealerbuilt.com returns a 403 Cloudflare challenge with zero visible words, zero JSON-LD schemas, zero headings, and no structured data. The site is effectively a blank wall.

   What to change: Serve actual HTML content with product descriptions, headings, and structured data without requiring a JS challenge.

6. Zero search results for brand and product names High

   DuckDuckGo returns zero results for queries including 'DealerBuilt', 'DealerBuilt DMS', 'dealerbuilt.com', and 'LightYear DMS'. No G2 or Capterra review pages appear in search results. The site has no discoverable subdomains or indexed URLs.

   What to change: Ensure the site is publicly accessible and indexed by search engines; build external backlinks and listings on review platforms.

7. No Wayback Machine snapshot exists Medium

   The Wayback Machine has no snapshot of dealerbuilt.com, indicating the site has never been publicly accessible or has been blocked from archiving.

   What to change: Allow the site to be crawled by the Internet Archive and other archivers.

8. OpenAI and Anthropic verification tokens present but ignored Medium

   DNS TXT records include openai-domain-verification and anthropic-domain-verification tokens, indicating proactive registration with both AI providers. However, both GPTBot and ClaudeBot are blocked at the Cloudflare layer, rendering the tokens useless.

   What to change: Configure Cloudflare to allow GPTBot and ClaudeBot access despite the verification tokens.

9. LLM cold knowledge contradicts site reality Medium

   The LLM's cold knowledge describes DealerBuilt as a DMS provider with LightYear DMS, CRM, digital retailing, and positive reviews. The actual site reveals none of this information, creating a total gap between what the LLM knows and what the site exposes.

   What to change: Publish product pages, pricing, about information, and customer testimonials on the site to align with LLM knowledge.

10. Zero JSON-LD schemas or structured data on homepage Medium

    The homepage contains no JSON-LD schemas, no structured data types, no FAQ, no comparison table, no list, and no definition patterns. No answer-format signals are present.

    What to change: Add JSON-LD structured data for organization, product, FAQ, and other relevant schemas.

11. No external mentions or backlinks found Medium

    Web searches for the brand and product names return zero results, indicating no external mentions, backlinks, or listings on review sites like G2 or Capterra.

    What to change: Build external backlinks and ensure the company is listed on relevant directories and review platforms.

What's working

• OpenAI and Anthropic domain verification tokens are configured — DNS TXT records include openai-domain-verification and anthropic-domain-verification tokens, showing proactive registration with both AI providers for crawler access.