lacarguy.com — AI Site Grade

Analysis

The Akamai Wall That Only Lets Claude Through

The site's live homepage returns HTTP 403 "Access Denied" to every browser and every major AI crawler — GPTBot, Google-Extended, PerplexityBot, OAI-SearchBot, Applebot-Extended, and ChatGPT-User all hit the same AkamaiGHost block. The sole exception is anthropic-ai, which receives a 200 with 309KB of full HTML served from an nginx backend behind the Akamai edge. This creates a bizarre inversion: the one AI crawler that can read the site is the one whose parent company explicitly prohibits using customer data for training, while every other model sees nothing but a 15-word error page.

Crawler Access

The robots.txt is itself blocked by Akamai for standard browsers (returns 403), but the version served to anthropic-ai reveals a detailed AI-crawler policy. GPTBot, OAI-SearchBot, ChatGPT-User, Claude-User, Claude-SearchBot, and PerplexityBot are all allowed to crawl the site but disallowed from accessing JS, CSS, JSON, API endpoints, and /pixall/ tracking paths. No AI bot is fully disallowed from the root. The llms.txt exists and is 880KB — an enormous file that appears to be an auto-generated dump of every page on the site, not a curated summary. The sitemap contains hundreds of URLs with lastmod dates as far out as 2026-05-30, suggesting the timestamps are synthetic or the site is pre-generating future content.

Cold-Knowledge Gap

LLM knowledge of lacarguy.com is zero. The model reports no verifiable information about the brand, its products, or its reputation. The site itself describes a 13-dealership family group serving Los Angeles since 1974, selling Audi, Genesis, Honda, Hyundai, Kia, Lexus, Porsche, Lotus, Subaru, Toyota, Volkswagen, and Fisker — a major multi-franchise operation with 900+ employees. The gap between this substantial real-world presence and the complete absence of cold LLM knowledge is total. No external search results, reviews, Reddit threads, or press mentions surfaced for the brand name or domain.

Schema Posture

The homepage HTML served to anthropic-ai contains no JSON-LD structured data in the visible snippet. The page relies on the Dealer.com (CDK Global) platform's default markup, which typically injects schema via JavaScript rather than server-side HTML. This means crawlers that execute JS (including most AI crawlers) may eventually see schema, but the initial HTML payload — which is what anthropic-ai and other text-preference crawlers consume — is schema-free. The heading structure is well-organized with H1 ("LAcarGUY Family of Dealerships"), H2s for each service category, and H3s for individual dealerships, but no FAQPage, Product, or AutoDealer schema was detected in the server-rendered HTML.

External Signals

The domain has zero indexed pages in search engines and zero external mentions across the open web. The Wayback Machine shows a single snapshot from a future date (March 2026), suggesting the site may have recently relaunched or changed platforms. The DNS points to a standard GoDaddy/Outlook setup with Akamai as the CDN. The complete absence of off-domain signals means AI engines have no secondary sources to triangulate the brand's identity, reputation, or inventory.

Findings

1. Akamai blocks all AI crawlers except anthropic-ai High

   The homepage returns HTTP 403 to every browser and major AI crawler (GPTBot, Google-Extended, PerplexityBot, OAI-SearchBot, Applebot-Extended, ChatGPT-User). Only anthropic-ai receives a 200 response with full HTML.

   What to change: Allow other major AI crawlers (GPTBot, Google-Extended, PerplexityBot, etc.) to access the site by adjusting Akamai WAF rules to permit their user-agent strings.

2. robots.txt blocked by Akamai for standard browsers High

   The robots.txt file returns 403 when accessed by a standard browser, preventing search engines and crawlers that don't mimic anthropic-ai from reading crawl directives.

   What to change: Ensure robots.txt is publicly accessible without authentication or IP restrictions.

3. Zero LLM knowledge of lacarguy.com High

   The model reports no verifiable information about the brand, its products, or its reputation. No external search results, reviews, or press mentions exist for the domain or brand name.

   What to change: Improve AI visibility by allowing more crawlers and building external signals through listings, reviews, and press.

4. No JSON-LD structured data in server-rendered HTML High

   The homepage HTML served to anthropic-ai contains no JSON-LD structured data. Schema is likely injected via JavaScript, which text-preference AI crawlers do not execute.

   What to change: Add server-side JSON-LD structured data (e.g., AutoDealer, Product, FAQPage) to all key pages so AI crawlers can parse it without JavaScript.

5. Zero indexed pages and external mentions High

   The domain has no indexed pages in search engines and no external mentions across the open web. The Wayback Machine shows only a single future-dated snapshot.

   What to change: Build external signals by claiming business listings, encouraging reviews, and generating press coverage.

6. llms.txt is an 880KB uncurated dump Medium

   The llms.txt file is 880KB and appears to be an auto-generated dump of every page on the site, not a curated summary of key resources.

   What to change: Replace the auto-generated dump with a curated llms.txt that lists only the most important pages and summaries.

7. Sitemap contains future lastmod dates Medium

   The sitemap includes hundreds of URLs with lastmod dates as far out as 2026-05-30, suggesting synthetic timestamps or pre-generated future content.

   What to change: Ensure lastmod dates reflect actual last modification times to avoid confusing crawlers.

8. robots.txt disallows JS, CSS, and API endpoints for AI bots Medium

   Allowed AI bots are disallowed from accessing JS, CSS, JSON, API endpoints, and /pixall/ tracking paths, which may prevent them from rendering pages fully.

   What to change: Allow AI bots to access JS and CSS files so they can render pages more completely, unless there are security concerns.

9. Homepage returns 403 to all browsers High

   The homepage returns HTTP 403 to standard browsers, making the site inaccessible to human visitors and search engine crawlers that mimic browsers.

   What to change: Allow browser user-agents to access the homepage by adjusting Akamai WAF rules.

What's working

• anthropic-ai crawler is allowed and receives full HTML — The anthropic-ai crawler receives a 200 response with 309KB of full HTML, enabling Claude to index the site's content.
• Detailed robots.txt with AI-specific directives — The robots.txt served to anthropic-ai contains specific rules for multiple AI crawlers, showing awareness of AI bot management.
• llms.txt file is present and accessible — The site hosts an llms.txt file that is accessible to anthropic-ai, providing a potential resource for AI crawlers.
• Sitemap is accessible and contains hundreds of URLs — The sitemap.xml is accessible to anthropic-ai and lists hundreds of URLs, providing a comprehensive map of the site.
• Well-organized heading structure on homepage — The homepage uses a clear hierarchy of H1, H2, and H3 tags, which helps crawlers understand page structure.
• Substantial real-world presence with 13 dealerships — The site describes a large multi-franchise dealership group with 900+ employees, indicating a significant business that could benefit from AI visibility.