shottenkirk.com — AI Site Grade

Analysis

GPTBot and Google-Extended get hard 404s on every page while all other AI crawlers receive full 200 responses with rich content — a selective, server-level block that is invisible to the robots.txt and likely unintentional.

Crawler Access

The robots.txt at shottenkirk.com contains no AI-bot directives whatsoever. It blocks MJ12bot, SemrushBot, and BLEXBot, then applies a generic * rule disallowing /wp-admin/ and /wp-includes/. No mention of GPTBot, Google-Extended, ClaudeBot, PerplexityBot, or any other AI crawler. Despite this permissive stance, compare_bot_access reveals that GPTBot and Google-Extended receive a hard 404 on both the homepage and vehicle detail pages, while ClaudeBot, anthropic-ai, OAI-SearchBot, ChatGPT-User, PerplexityBot, Perplexity-User, Bytespider, and Applebot-Extended all get full 200 responses with identical byte payloads to a browser. This is a server-level (nginx) UA-based block, not a robots.txt directive — and it is inconsistent: OpenAI's OAI-SearchBot passes while GPTBot fails, suggesting a misconfigured rule rather than a deliberate policy. No llms.txt exists (404).

Cold-Knowledge Gap

The LLM's prior knowledge describes Shottenkirk as a "Midwest" group operating in "Illinois, Indiana, Iowa, and Missouri" with a "No Haggle" pricing model. The actual site reveals a much larger footprint: 28 locations across 7 states (adding California, Texas, Georgia, Alabama, and Florida to the Midwest core). The "No Haggle" claim — a notable differentiator in the cold knowledge — is entirely absent from the homepage, about page, and vehicle detail pages. The site positions itself on "family approach" and "customer first" but never mentions the pricing model that the LLM considers a signature trait. The group's founding story (Dean and Bob's Chevrolet in Abingdon, IL, 60+ years ago) is present on the about page but the LLM had no detail on it.

Schema Posture

Every page carries AutoDealer, Organization, and WebSite schema with SearchAction — a solid baseline. However, the AutoDealer schema is critically degraded: the address fields (streetAddress, addressLocality, addressRegion, postalCode) are all empty strings, and the telephone number is 555.555.5555 — a placeholder. The openingHoursSpecification array contains 10 overlapping entries with impossible values (e.g., Sunday opens at 00:00 and closes at 12:00 repeated across multiple departments). Vehicle detail pages include Car and Product schema with rich feature lists, but the group-level schema is polluted with placeholder data that undermines local-search and knowledge-graph extraction.

Content & External Signals

The homepage is a text-rich, server-rendered page (~1,200 words) with clear heading hierarchy and comparison language. No FAQ schema or FAQ-style content exists anywhere. The "In the Community" page is extremely thin (39 words of visible text) — mostly image galleries with no alt text or descriptive content. The blog/post sitemap contains only 6 posts, all from June 2024, indicating content production has stopped. The page sitemap shows blog-style articles last updated in March-May 2024 — stale. Web searches for the brand returned zero results, suggesting very low external citation volume and minimal off-domain signals for AI engines to reference.

Findings

1. GPTBot and Google-Extended receive hard 404s on all pages High

   GPTBot and Google-Extended get a hard 404 on both the homepage and vehicle detail pages, while all other AI crawlers (ClaudeBot, OAI-SearchBot, PerplexityBot, etc.) receive full 200 responses. This server-level nginx block is inconsistent and likely unintentional, as OAI-SearchBot passes while GPTBot fails.

   What to change: Remove the server-level UA-based block that returns 404 for GPTBot and Google-Extended, or update it to allow these bots to access the site.

2. AutoDealer schema contains empty address fields and placeholder phone number High

   The AutoDealer schema on every page has empty strings for streetAddress, addressLocality, addressRegion, and postalCode, and the telephone number is 555.555.5555. This placeholder data degrades local-search and knowledge-graph extraction.

   What to change: Populate the AutoDealer schema with accurate address and phone data for each location, and remove placeholder values.

3. AutoDealer openingHoursSpecification contains overlapping and impossible values Medium

   The openingHoursSpecification array has 10 overlapping entries with impossible values, such as Sunday opening at 00:00 and closing at 12:00 repeated across multiple departments. This pollutes the schema and may confuse knowledge graphs.

   What to change: Clean the openingHoursSpecification to contain only valid, non-overlapping hours for each department.

4. No llms.txt file exists Medium

   The site does not provide an llms.txt file, which is a recommended way to guide AI crawlers to key content. Its absence means AI engines lack a structured entry point.

   What to change: Create an llms.txt file listing important pages like locations, about, and vehicle inventory.

5. No-haggle pricing model not mentioned on site despite being a key differentiator Medium

   The LLM's prior knowledge identifies Shottenkirk as a 'No Haggle' dealership, but this pricing model is entirely absent from the homepage, about page, and vehicle detail pages. The site emphasizes 'family approach' and 'customer first' but omits the signature pricing policy.

   What to change: Add prominent mention of the 'No Haggle' pricing model on the homepage and about page.

6. LLM prior knowledge underestimates geographic footprint Medium

   The LLM's prior knowledge describes Shottenkirk as a Midwest group with locations in 4 states, but the actual site shows 28 locations across 7 states including California, Texas, Georgia, Alabama, and Florida. This mismatch may cause AI engines to provide incomplete information.

   What to change: Ensure location pages and schema accurately reflect the full geographic footprint to update AI knowledge.

7. In the Community page has very thin text content Low

   The 'In the Community' page contains only 39 words of visible text, mostly image galleries with no alt text or descriptive content. This provides little substance for AI crawlers to index.

   What to change: Add descriptive text and alt attributes to images on the In the Community page.

8. Blog content production has stopped since June 2024 Low

   The blog/post sitemap contains only 6 posts, all from June 2024, and the page sitemap shows blog-style articles last updated in March-May 2024. This indicates content production has ceased, reducing fresh signals for AI engines.

   What to change: Resume regular blog posting to provide fresh content for AI crawlers.

9. No FAQ schema or FAQ-style content exists Low

   The site does not use FAQ schema or present FAQ-style content, missing an opportunity to appear in AI-generated answers and rich snippets.

   What to change: Add FAQ schema to pages that answer common customer questions.

10. Very low external citation volume for the brand Medium

    Web searches for 'Shottenkirk Automotive Group' and related queries returned zero results, indicating minimal off-domain signals for AI engines to reference.

    What to change: Invest in local SEO, press releases, and partnerships to increase external citations.

What's working

• Robots.txt does not block any AI crawlers — The robots.txt file contains no directives targeting AI bots, allowing most AI crawlers to access the site freely.
• Most AI crawlers receive full 200 responses with rich content — ClaudeBot, OAI-SearchBot, PerplexityBot, and others get full HTML content identical to a browser, ensuring their access to site information.
• Every page includes AutoDealer, Organization, and WebSite schema with SearchAction — The site has a solid baseline of structured data including AutoDealer, Organization, and WebSite schema with SearchAction, which helps search engines understand the site.
• Vehicle detail pages include Car and Product schema with rich feature lists — Vehicle detail pages contain Car and Product schema with detailed feature lists, providing structured data for AI engines to extract vehicle information.
• Homepage is text-rich with clear heading hierarchy — The homepage contains approximately 1,200 words of server-rendered text with a clear heading hierarchy, providing substantial content for AI crawlers.
• Sitemap exists and contains 80 URLs — The sitemap at sitemap.xml is accessible and lists 80 URLs, helping crawlers discover site content.
• About page includes founding story and history — The about page provides a detailed founding story (Dean and Bob's Chevrolet in Abingdon, IL, 60+ years ago), which can help AI engines build a narrative.