caronsale.com — AI Site Grade

Analysis

CarOnSale's site is fully open to every major AI crawler, but the cold-knowledge gap reveals a brand that AI models describe as a "challenger" while the site positions itself as a market leader with a Mercedes-Benz partnership and a fresh EUR 70M Series C — facts the model knows nothing about.

Crawler Access

Every major AI bot — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Applebot-Extended, anthropic-ai — receives a full 200 response with identical byte payload (257 KB) to a browser visit. The sole exception is Bytespider (ByteDance), which gets a 403 from Cloudflare. The robots.txt is a single Allow: / for User-Agent: * with no AI-specific rules at all — no GPTBot block, no ClaudeBot restriction, no crawl-delay. The site runs on Cloudflare (DNS via Cloudflare nameservers, server header cloudflare, strict-transport-security enabled). No llms.txt exists (404s to a Webflow 404 page). The sitemap.xml contains 410 URLs covering 18+ language variants, brand-model pages, glossary entries, and regional landing pages.

Cold-Knowledge Gap

When queried cold, the LLM describes CarOnSale as a "challenger to traditional wholesale channels" founded in 2015 in Berlin, with mixed Trustpilot reviews and venture funding from IBB Ventures. The actual site tells a dramatically different story: founded in 2018 (not 2015), the company has raised a EUR 70M Series C led by Northzone (announced July 2025), is the official Mercedes-Benz partner for marketing lease returns across Europe, acquired Alpha Online in 2024, and claims ~2,500 locations, ~40,000 active users, ~100,000 vehicles auctioned annually. The model's prior is 2-3 years stale — it knows nothing about the Mercedes deal, the Series C, the Alpha Online acquisition, or the scale metrics the site prominently displays.

Schema Posture

The homepage carries a single WebPage schema with embedded Organization data including legal name (Castle Tech GmbH), address, contact info, and sameAs links. The /buy page repeats this pattern. The Mercedes A-Class model page is the richest — it includes BreadcrumbList, VehicleModel with brand, and datePublished/dateModified. However, critical pages are schema-bare: the /about-us, /sell, /blog, /press, and /autos/all pages contain zero JSON-LD. No FAQPage schema is used despite FAQ sections on multiple pages. No Product or Auction schema is applied to the thousands of vehicle listings the site surfaces.

External Signals

The press room documents coverage from Startbase, Autohaus.de, Automobilwoche, kfz-betrieb, Silicon Canals, Tech Funding News, VC Magazin — all covering the Mercedes-Benz partnership and Series C. The site links to Datenschutzexperte.de privacy seal, Greenhouse job board, and app store listings. No Trustpilot or Google review widgets appear on the site, and the cold model's mention of mixed reviews cannot be verified from the site itself. The footer copyright reads 2026, suggesting the site may be publishing content dated far into the future.

Surprising Details

The llms.txt 404 returns a full Webflow HTML page with a canonical pointing to /en/404 — the site is running on Webflow (cdn.prod.website-files.com) with Intellimize A/B testing and Weglot for 18-language translation. The homepage inLanguage is set to "de" even on the English URL. The blog has not published since at least 2022 based on visible post titles. The site claims "1968 vehicles" on the homepage but "2918 vehicles" on the buy page — a discrepancy suggesting stale or dynamic count mismatches. The copyright year "2026" appears in the footer while the current year is 2025, indicating a forward-dated publishing setup.

Findings

1. Cold-knowledge gap: AI models lack key brand facts High

   When queried cold, an LLM describes CarOnSale as a 2015-founded challenger with mixed reviews, missing the 2018 founding, EUR 70M Series C, Mercedes-Benz partnership, Alpha Online acquisition, and scale metrics (~2,500 locations, ~40,000 active users, ~100,000 vehicles auctioned annually).

   What to change: Publish structured data (Organization, Event, Product) on key pages and ensure high-value facts appear in crawlable text. Consider creating an llms.txt file with authoritative brand information.

2. Missing schema on key pages High

   The /about-us, /sell, /blog, /press, and /autos/all pages contain zero JSON-LD. No FAQPage, Product, or Auction schema is used despite FAQ sections and thousands of vehicle listings.

   What to change: Add JSON-LD schemas (Organization, FAQPage, Product, Auction) to all relevant pages, especially vehicle listing pages and the press room.

3. No llms.txt file Medium

   The llms.txt file returns a 404 (Webflow page), missing an opportunity to provide AI crawlers with a curated summary of the site's content and key facts.

   What to change: Create an llms.txt file at the root with a concise overview of the company, key pages, and structured data hints.

4. Homepage language set to German on English URL Medium

   The homepage's inLanguage attribute is set to 'de' even on the English URL (https://www.caronsale.com/en), which may confuse crawlers about the page's language.

   What to change: Set the inLanguage attribute to 'en' on English-language pages.

5. Vehicle count discrepancy between homepage and buy page Medium

   The homepage claims '1968 vehicles' while the buy page shows '2918 vehicles', a 48% difference that may indicate stale or mismatched counts.

   What to change: Ensure vehicle counts are consistent across pages or use a single dynamic source.

6. Footer copyright year set to 2026 Low

   The footer displays '© 2026 CarOnSale' while the current year is 2025, which may appear as an error or future-dated content.

   What to change: Update the copyright year to the current year or use a dynamic year.

7. Blog appears inactive since 2022 Low

   The blog page shows no recent posts, with visible titles suggesting no publication since at least 2022, which may signal outdated content to crawlers.

   What to change: Either resume publishing or remove the blog section to avoid appearing stale.

8. Bytespider (ByteDance) blocked by Cloudflare Low

   Bytespider receives a 403 response, preventing ByteDance's AI crawler from accessing the site.

   What to change: If ByteDance's AI products are a target audience, allow Bytespider access.

What's working

• Full access for all major AI crawlers — Every major AI bot (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, etc.) receives a full 200 response with identical content to a browser visit, with no blocks or restrictions.
• Comprehensive sitemap with 410 URLs — The sitemap.xml contains 410 URLs covering 18+ language variants, brand-model pages, glossary entries, and regional landing pages, aiding crawler discovery.
• Rich schema on Mercedes A-Class model page — The model page includes BreadcrumbList, VehicleModel, and datePublished/dateModified schemas, providing structured data for a key product page.
• Press room with media coverage links — The press page documents coverage from reputable outlets (Startbase, Autohaus.de, Automobilwoche, etc.) covering the Mercedes-Benz partnership and Series C, providing external validation.
• Organization schema on homepage — The homepage includes WebPage schema with embedded Organization data (legal name, address, contact info, sameAs links), providing basic entity information.