irwinzone.com — AI Site Grade

Analysis

I have enough data to write a thorough audit. Let me compile the findings.

Cold-Knowledge Brand Confusion: LLM Believes Irwin Sells GM, Site Sells Toyota/Ford/Hyundai/Chevy/Lincoln

The most significant AI-visibility gap for irwinzone.com is a fundamental brand-positioning failure in what large language models know about the dealership group. When queried cold, a frontier LLM described Irwin Automotive Group as selling "Chevrolet, Buick, GMC, and Cadillac" — a pure GM portfolio. The actual site sells Ford, Toyota, Chevrolet, Hyundai, and Lincoln. The LLM's prior is entirely wrong about the brand's core identity, missing three of the five actual brands (Toyota, Ford, Hyundai) and inventing two that are not sold (Buick, GMC).

Crawler Access

All major AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Applebot-Extended, Bytespider, anthropic-ai, ChatGPT-User — receive a full 200 response with identical byte payload (~438KB) to a browser baseline. No UA-based blocking exists. The site runs on nginx behind Varnish CDN, hosted via DealerOn (DNS: dns1.dealeron.com, dns2.dealeron.com). The robots.txt contains zero AI-bot-specific rules — a single User-Agent: * block with crawl-delay 10 and standard disallows for AJAX/print/rebate endpoints. No GPTBot, ClaudeBot, or PerplexityBot directives exist. The /llms.txt returns 404 (serves a 142KB HTML error page). The sitemap.xml is present with 1,533 URLs indexed.

Cold-Knowledge Gap

The LLM cold-knowledge output is the audit's central finding. It describes Irwin as "a family-owned dealership group primarily selling new and used vehicles, with a focus on Chevrolet, Buick, GMC, and Cadillac." The actual site prominently features Ford, Toyota, Chevrolet, Hyundai, and Lincoln as new-car brands, with dedicated search pages for each (/searchnew/new-ford.html, /searchnew/new-toyota.html, etc.). The site has no Buick, GMC, or Cadillac inventory or pages. The LLM also missed the group's explicit claims: "#1 in NH Google Reviews" and "#1 in NH Cars.com Reviews" with 12,500+ Google reviews and a 4.7-star rating — strong social-proof signals the model did not surface.

Schema Posture

The site uses AutomotiveBusiness JSON-LD with AutoDealer, AutoRepair, and AutoPartsStore sub-departments. The schema includes address, geo coordinates, telephone, opening hours, and sameAs links to Twitter, Facebook, YouTube, and Cars.com. However, the description field is generic boilerplate: *"Irwin Automotive Group is a Group dealer in Laconia, NH. We specialize in new Group, used vehicles, service, and financing."* — the word "Group" appears as a placeholder rather than listing actual brands. No FAQPage, Product, Vehicle, or Review schema is present despite the site having comparison content (e.g., "Toyota RAV4 vs. Highlander") and thousands of reviews.

Content Freshness & Stale Pages

The sitemap contains a URL /new-2021-ford-f-150-laconia-nh — a page promoting the 2021 Ford F-150 as "new." This page remains indexed and live in 2025, describing a vehicle now four model years old as "new." The blog post "What's Going on in the Used Car Market" references market conditions from a period of "fluctuating prices" without a publication date. The Wayback Machine shows a snapshot from December 2025, confirming the homepage is actively crawled, but the stale model-year content undermines freshness signals for AI engines that prioritize recency.

External Signals

The site links to brand-specific subdomains (irwinzonetoyota.com, irwinzoneford.com, irwinzonechevy.com, irwinhyundai.com, irwinzonelincoln.com) for specials — a fragmented domain strategy that splits brand authority across five separate domains rather than consolidating under irwinzone.com. The sameAs schema references @IrwinToyotaFord on Twitter, which reinforces the Toyota/Ford identity the LLM missed. No FAQ, comparison table, or structured review markup exists on any page examined, despite the site claiming 12,000+ reviews and offering vehicle comparisons.

Findings

1. LLM cold-knowledge misidentifies Irwin as a GM-only dealership High

   When queried cold, a frontier LLM described Irwin Automotive Group as selling Chevrolet, Buick, GMC, and Cadillac. The actual site sells Ford, Toyota, Chevrolet, Hyundai, and Lincoln. The LLM's prior is entirely wrong about the brand's core identity, missing three of the five actual brands and inventing two that are not sold.

   What to change: Add explicit brand listings in structured data (e.g., AutomotiveBusiness schema with makesAccepted) and publish an llms.txt file that lists the actual brands sold. Ensure homepage and key pages prominently list all brands in plain text.

2. Missing /llms.txt file Medium

   The site returns a 404 for /llms.txt, serving a 142KB HTML error page instead. This file is a standard way to provide AI crawlers with a concise summary of the site's content and key facts.

   What to change: Create an /llms.txt file that lists the dealership's actual brands, location, services, and key differentiators (e.g., #1 in NH Google Reviews).

3. Stale 'new' vehicle page for 2021 Ford F-150 Medium

   The sitemap contains a URL /new-2021-ford-f-150-laconia-nh that promotes the 2021 Ford F-150 as 'new' in 2025. This page remains indexed and live, describing a vehicle now four model years old as new, undermining freshness signals for AI engines.

   What to change: Update or remove the page for the 2021 Ford F-150. Replace with current model year content or redirect to a generic new Ford F-150 page.

4. Schema description uses placeholder 'Group' instead of actual brands Medium

   The AutomotiveBusiness JSON-LD description reads: 'Irwin Automotive Group is a Group dealer in Laconia, NH. We specialize in new Group, used vehicles, service, and financing.' The word 'Group' appears as a placeholder rather than listing actual brands like Ford, Toyota, Chevrolet, Hyundai, and Lincoln.

   What to change: Update the schema description to list the actual brands sold (Ford, Toyota, Chevrolet, Hyundai, Lincoln) and remove the placeholder 'Group' text.

5. No Vehicle or Product schema on vehicle pages Medium

   Despite having vehicle comparison content (e.g., 'Toyota RAV4 vs. Highlander') and inventory pages, the site does not use Vehicle or Product structured data. This limits the ability of AI engines to extract specific vehicle details.

   What to change: Add Vehicle structured data (e.g., Vehicle, Car) to inventory and comparison pages, including make, model, year, trim, price, and availability.

6. No Review or AggregateRating schema despite 12,000+ reviews Medium

   The site claims 12,500+ Google reviews and a 4.7-star rating, but no Review or AggregateRating structured data is present on any examined page. This prevents AI engines from surfacing these strong social-proof signals.

   What to change: Add AggregateRating schema to the homepage and testimonials page, referencing the Google review count and average rating.

7. Brand authority split across five separate domains Medium

   The site links to brand-specific subdomains (irwinzonetoyota.com, irwinzoneford.com, irwinzonechevy.com, irwinhyundai.com, irwinzonelincoln.com) for specials, fragmenting brand authority and link equity away from the main irwinzone.com domain.

   What to change: Consolidate brand-specific content under irwinzone.com subdirectories (e.g., /ford/, /toyota/) rather than separate domains, to centralize authority and improve AI understanding of the full brand lineup.

8. No FAQPage schema on comparison or FAQ content Low

   The site has comparison articles (e.g., 'Toyota RAV4 vs. Highlander') that could benefit from FAQPage schema, but none is present. This reduces the chance of appearing in AI-generated answers.

   What to change: Add FAQPage schema to comparison articles and any page that answers common customer questions.

9. Blog posts lack visible publication dates Low

   The blog post 'What's Going on in the Used Car Market' does not display a publication date, making it harder for AI crawlers to assess content freshness.

   What to change: Add visible publication dates and dateModified metadata to all blog posts and articles.

10. Robots.txt has no AI-bot-specific directives Low

    The robots.txt file contains only a generic User-Agent: * block with crawl-delay 10 and standard disallows. No directives for GPTBot, ClaudeBot, PerplexityBot, or other AI crawlers exist, which is a missed opportunity to guide AI crawler behavior.

    What to change: Add specific directives for AI crawlers (e.g., Allow for GPTBot, ClaudeBot) to ensure they can access key content while respecting crawl-delay.

What's working

• All major AI crawlers receive full 200 response — All 11 tested AI crawlers (GPTBot, ClaudeBot, PerplexityBot, etc.) receive a full 200 response with identical content to a browser, with no UA-based blocking.
• AutomotiveBusiness JSON-LD with sub-departments — The site uses AutomotiveBusiness schema with AutoDealer, AutoRepair, and AutoPartsStore sub-departments, including address, geo coordinates, telephone, opening hours, and sameAs links.
• Sitemap.xml present with 1,533 URLs — The sitemap.xml is accessible and contains 1,533 URLs, providing a comprehensive index for crawlers.
• Site prominently claims #1 in NH Google Reviews and 12,500+ reviews — The homepage and testimonials page highlight strong social proof: #1 in NH Google Reviews, #1 in NH Cars.com Reviews, 12,500+ Google reviews, and a 4.7-star rating.
• Vehicle comparison articles provide rich content — The site publishes comparison articles like 'Toyota RAV4 vs. Highlander' that offer detailed, useful content for AI crawlers to index.
• Homepage actively crawled by Wayback Machine — The Wayback Machine has a recent snapshot from December 2025, indicating the homepage is regularly archived.