rafihautogroup.com — AI Site Grade

Analysis

The sitemap and all non-homepage pages time out for crawlers

The homepage loads fully for every AI bot tested, but every deeper page — /about/, /locations/, /blogs/, /careers/, /press-release/, /2026-vs-2025-jeep-grand-cherokee-comparison/ — returns a read timeout. The sitemap_index.xml also times out. This means AI crawlers can see the homepage but cannot discover or access the rest of the site's content.

Crawler Access

All 11 AI bot user-agents tested (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Bytespider, Applebot-Extended, anthropic-ai, Perplexity-User, and a browser baseline) receive a 200 status with identical byte size (491,468 bytes) from the homepage. No UA-based blocking occurs. The site runs on Flywheel/5.1.0 hosting with Fastly CDN caching. The robots.txt has no AI-bot-specific rules — only a catch-all User-agent: * disallowing /*? and calendar/event action paths, with a Crawl-delay: 3. No GPTBot, ClaudeBot, or other AI crawler is mentioned. The llms.txt exists (generated by Yoast SEO v27.6) and lists pages, posts, job openings, and categories — but the URLs it references are the same ones that time out.

Cold-Knowledge Gap

A frontier LLM queried cold about Rafih Auto Group describes it as a Canadian dealership group based in the GTA serving Ontario with Ford, Lincoln, Mazda, and Kia brands, family-owned for over 30 years. The actual site reveals a far larger operation: 26+ dealerships across Ontario, Ohio, and Michigan representing 30+ brands including Mercedes-Benz, Porsche, BMW, Lexus, Jaguar, Land Rover, Audi, and MINI. The cold knowledge misses the entire US presence, the luxury brand portfolio, the scale (26 locations vs. implied few), and the founder Terry Rafih. The gap between what AI models know and what the site actually contains is extreme — the model describes a small regional chain when the site describes a binational luxury auto group.

Schema Posture

The homepage includes a well-structured JSON-LD schema block with WebPage, WebSite, Organization, ImageObject, and BreadcrumbList types. The Organization schema includes name, URL, and logo. However, there is no LocalBusiness or AutoDealer schema despite the site representing 26+ physical dealership locations. No Product schema for vehicle inventory. No FAQPage schema despite comparison content on the site. The dateModified field reads 2026-02-03 — a future date — suggesting a plugin or theme bug.

External Signals

Web searches for "Rafih Auto Group" across multiple query variations returned zero indexed results from DuckDuckGo. The US domain (rafihautogroup.us) returns a 403 Forbidden from Cloudflare for all bots and browsers — the entire US-facing site is inaccessible to crawlers. The inventory subdomain (inventory.rafihautogroup.com) also returns a Cloudflare 403 block. No Wayback Machine snapshot exists for the domain. The site has essentially no discoverable external footprint — no reviews, no press mentions, no Reddit threads surfaced in search.

Findings

1. Sitemap and deep pages time out — sitemap_index.xml, /about/, /locations/, /blogs/, /careers/, /press-release/, and comparison articles all fail with read timeouts. AI crawlers can only reliably access the homepage. 2. No AI-bot directives in robots.txt — No mention of GPTBot, ClaudeBot, Google-Extended, or any other AI crawler. The catch-all rule disallows /*? which could block some parameterized URLs. 3. Cold LLM knowledge is severely outdated — The model describes a small GTA-based Ford/Mazda/Kia dealer. The site describes 26+ locations across Canada and the US with Mercedes-Benz, Porsche, BMW, Lexus, Audi, Jaguar, Land Rover, and MINI. 4. US domain completely blocked — rafihautogroup.us returns 403 to all bots and browsers via Cloudflare. The entire US dealership network is invisible to AI crawlers. 5. Inventory subdomain blocked — inventory.rafihautogroup.com returns Cloudflare 403. Vehicle inventory is not crawlable. 6. No AutoDealer or LocalBusiness schema — Despite being a multi-location auto group, the homepage schema uses only generic Organization type. No per-location schemas, no Product schemas for vehicles. 7. Future date in schema — dateModified is set to 2026-02-03, which may confuse crawlers about content freshness. 8. Zero external search footprint — Multiple search queries returned no results for the brand name, domain, or founder. No reviews, press, or social mentions surfaced. 9. No Wayback Machine history — No snapshots exist, suggesting the domain may be relatively new or previously unindexed. 10. llms.txt exists but references unreachable URLs — Yoast-generated llms.txt lists pages that time out when fetched, providing no value to AI crawlers.

Findings

1. Sitemap and all non-homepage pages time out for crawlers High

   The sitemap_index.xml and every deeper page (/about/, /locations/, /blogs/, /careers/, /press-release/, /2026-vs-2025-jeep-grand-cherokee-comparison/) return read timeouts. AI crawlers can only reliably access the homepage.

   What to change: Investigate server configuration or CDN rules causing timeouts on non-homepage URLs. Ensure the sitemap and all content pages return 200 within a reasonable timeout for crawlers.

2. US domain returns 403 Forbidden to all bots and browsers High

   The US-facing domain rafihautogroup.us returns a 403 Forbidden from Cloudflare for all 11 tested user-agents, making the entire US dealership network invisible to AI crawlers.

   What to change: Remove the Cloudflare block for legitimate AI crawlers on rafihautogroup.us, or redirect the domain to the main site.

3. Inventory subdomain blocked by Cloudflare 403 High

   The inventory subdomain inventory.rafihautogroup.com returns a Cloudflare 403 block, preventing crawlers from accessing vehicle inventory.

   What to change: Allow AI crawlers through Cloudflare on the inventory subdomain, or ensure inventory is accessible via the main domain.

4. Cold LLM knowledge severely underestimates dealership scale and brands High

   A frontier LLM queried cold describes Rafih Auto Group as a small GTA-based Ford/Mazda/Kia dealer, missing the actual 26+ locations across Canada and the US, luxury brands (Mercedes-Benz, Porsche, BMW, Lexus, Audi, Jaguar, Land Rover, MINI), and founder Terry Rafih.

   What to change: Ensure all dealership locations, brands, and key personnel are prominently listed on crawlable pages with structured data to improve AI knowledge.

5. No AutoDealer or LocalBusiness schema on homepage Medium

   The homepage JSON-LD uses only generic Organization type. No LocalBusiness or AutoDealer schema is present despite 26+ physical dealership locations. No Product schema for vehicle inventory.

   What to change: Add LocalBusiness or AutoDealer schema for each dealership location, and Product schema for vehicle inventory pages.

6. Schema dateModified set to future date 2026-02-03 Medium

   The homepage schema includes a dateModified value of 2026-02-03, which may confuse crawlers about content freshness and could be seen as a signal of low quality.

   What to change: Fix the dateModified field to reflect the actual last-modified date of the page.

7. Zero external search footprint for brand and domain Medium

   Multiple web searches for 'Rafih Auto Group', the domain, and founder returned zero results on DuckDuckGo. No reviews, press mentions, or social references surfaced.

   What to change: Build external signals through press releases, local business listings, social media profiles, and review platforms to improve discoverability.

8. No Wayback Machine snapshots exist for the domain Low

   The Wayback Machine has no snapshots for rafihautogroup.com, suggesting the domain may be relatively new or previously unindexed.

9. llms.txt references URLs that time out Medium

   The Yoast-generated llms.txt lists pages, posts, and job openings, but all referenced URLs time out when fetched, providing no value to AI crawlers.

   What to change: Ensure the URLs listed in llms.txt are accessible and return 200 status codes.

10. No AI-bot-specific directives in robots.txt Low

    The robots.txt does not mention GPTBot, ClaudeBot, Google-Extended, or any other AI crawler. The catch-all rule disallows /*? which could block some parameterized URLs.

    What to change: Add explicit allow rules for AI crawlers if they are not already allowed, and review the /*? disallow rule.

What's working

• Homepage returns 200 for all 11 tested AI bots — The homepage loads successfully for all tested AI crawlers with no UA-based blocking, ensuring the front door is open.
• llms.txt file is published and lists site content — The site has a Yoast-generated llms.txt that lists pages, posts, job openings, and categories, providing a structured entry point for AI crawlers.
• Homepage includes well-structured JSON-LD schema — The homepage contains JSON-LD with WebPage, WebSite, Organization, ImageObject, and BreadcrumbList types, providing basic structured data.
• robots.txt does not block any AI crawlers — The robots.txt has no AI-bot-specific disallow rules, meaning AI crawlers are not explicitly blocked from any paths.