pfannenberg.com — AI Site Grade

Analysis

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

The most striking finding: the homepage and all root-domain product pages are JS-rendered language selectors with only 32 words of visible text, while the /en/ subdirectory has rich content. ClaudeBot and Bytespider are blocked entirely (403). No llms.txt. Only WebSite schema everywhere — no Product, Organization, FAQPage, or BreadcrumbList schema on any page checked.

Let me write the audit.

Pfannenberg.com — AI-Visibility Audit

The homepage and every root-domain product page (e.g. /thermal-management/cooling-units/...) serve a 32-word JavaScript shell — a language selector with zero product content — while the /en/ subdirectory pages render full HTML with 500+ words. AI crawlers hitting the root domain see nothing useful; those hitting /en/ paths see rich content, but ClaudeBot and Bytespider are blocked with a 403 on all pages, while GPTBot, Google-Extended, OAI-SearchBot, PerplexityBot, and ChatGPT-User all pass through cleanly.

Crawler Access & Infrastructure

The robots.txt contains no AI-bot-specific directives — only a catch-all User-agent: * with Allow: / and disallows for /typo3/, /vendor/, contact pages, and login areas. The site runs on Apache with no CDN or WAF visible (no Cloudflare, no Akamai). The llms.txt returns a 404 (TYPO3 CMS error page). The sitemap.xml lists 1,763 URLs and is well-formed. The products.pfannenberg.com subdomain hosts a separate customer portal with richer product data but zero JSON-LD schema on any page there.

Content & Schema Posture

Every page on the main domain — homepage, about, product category, and individual product pages — injects only a single WebSite schema with name and URL. No Organization, Product, BreadcrumbList, FAQPage, or HowTo schema exists anywhere. The /en/ homepage is titled "F-gas compliance" and leads with 32 H2 news items about F-gas regulation, Green Series cooling units, IO-Link signaling, and sustainability — strong topical content that AI engines could use, but no structured data tells them what Pfannenberg actually sells. Product detail pages (e.g. DTI 9021) contain technical tables, download links, and benefit lists but lack Product schema with SKU, price, or specification properties.

Cold-Knowledge Gap

The LLM knows Pfannenberg as a "German-based company founded in 1954" specializing in thermal management and signaling, with "Premier" and "Pyra" product series. The actual site never mentions "Premier" — that series name appears to be stale training data. The site heavily promotes "Green Series," "X-Series," "Patrol," "Protect," and "Quadro" series, plus the new "Pfire" fire-alarm line and "One Pfannenberg" global initiative. The cold model knows nothing about the F-gas compliance positioning, which is the homepage's headline message. The model also misses the 10-year warranty, the PSS sizing software, the 3D-Coverage calculator, and the Kununu Top Company 2026 status — all prominently featured on the site.

External Signals

External search results for Pfannenberg return almost nothing — no Reddit threads, no review sites, no press coverage surfaced through standard queries. The site links to LinkedIn (881845), YouTube, and Xing, but these are not indexed by general web search in a way that shapes AI knowledge. The Kununu score (4.2/5, "Top Company 2026") is mentioned on the homepage but has no external citation or schema markup. The "What customers say about us" page exists at /en/about/what-customers-say-about-us/ but was not fetched — its mere existence suggests testimonial content that could be schema-marked as Review but almost certainly is not.

Surprising Contradictions

The homepage title "F-gas compliance" leads to a 404 when visiting /en/f-gas-compliance/ directly — the page does not exist as a standalone URL. The F-gas content is scattered across news items and product descriptions rather than living on a dedicated landing page. The root domain (pfannenberg.com) and the /en/ subdirectory serve fundamentally different content to crawlers: the root is a JS shell, the /en/ path is server-rendered HTML. Any AI crawler that starts at the root (as many do) gets 32 words and leaves. The products. subdomain has richer product data but zero schema and no cross-linking in the sitemap.

Findings

1. Root domain pages serve empty JavaScript shells to crawlers High

   The homepage and all root-domain product pages (e.g., /thermal-management/cooling-units/...) render only a 32-word language selector with no product content. AI crawlers hitting the root domain see nothing useful, while /en/ subdirectory pages have rich HTML content.

   What to change: Implement server-side rendering or static HTML for all root-domain pages so that crawlers receive full content regardless of JavaScript execution.

2. ClaudeBot and Bytespider receive 403 Forbidden on all pages High

   ClaudeBot and Bytespider are blocked with a 403 error on all pages, while GPTBot, Google-Extended, OAI-SearchBot, PerplexityBot, and ChatGPT-User pass through cleanly. This reduces AI visibility from two major crawlers.

   What to change: Remove the 403 block for ClaudeBot and Bytespider, or ensure they are allowed via server configuration.

3. No Product or Organization schema on any page High

   Every page on the main domain injects only a single WebSite schema with name and URL. No Product, Organization, BreadcrumbList, FAQPage, or HowTo schema exists anywhere, even on product detail pages with technical specifications.

   What to change: Add Product schema with SKU, price, and specifications to product pages, and Organization schema to the homepage and about page.

4. llms.txt returns 404 error Medium

   The llms.txt file at pfannenberg.com/llms.txt returns a 404 error (TYPO3 CMS error page), meaning AI crawlers cannot discover a curated set of important URLs.

   What to change: Create an llms.txt file listing key pages such as product categories, about page, and service pages.

5. LLM knowledge of Pfannenberg is outdated and incomplete Medium

   The LLM knows Pfannenberg as a German company founded in 1954 with 'Premier' and 'Pyra' series, but the site never mentions 'Premier'. The site promotes 'Green Series', 'X-Series', 'Patrol', 'Protect', 'Quadro', and 'Pfire' fire-alarm line, plus the 'One Pfannenberg' initiative. The F-gas compliance positioning, 10-year warranty, PSS sizing software, 3D-Coverage calculator, and Kununu Top Company 2026 status are all missing from the model's knowledge.

   What to change: Ensure key brand terms and product series are prominently mentioned in visible text and structured data to update LLM knowledge.

6. Near-zero external web presence on review and social platforms Medium

   Web searches for Pfannenberg on Reddit, LinkedIn, Kununu, and general review sites return zero results. The site links to LinkedIn, YouTube, and Xing, but these are not indexed in a way that shapes AI knowledge.

   What to change: Encourage customer reviews on third-party platforms and ensure social media profiles are publicly indexed.

7. F-gas compliance page returns 404 Medium

   The URL /en/f-gas-compliance/ returns a 404 error, even though the homepage title is 'F-gas compliance' and F-gas content is scattered across news items. There is no dedicated landing page for this key topic.

   What to change: Create a dedicated F-gas compliance landing page with comprehensive content and internal links.

8. Products subdomain lacks any structured data Medium

   The products.pfannenberg.com subdomain hosts richer product data but has zero JSON-LD schema on any page, missing opportunities for AI crawlers to understand product details.

   What to change: Add Product and Organization schema to all pages on the products subdomain.

9. BreadcrumbList schema missing on all pages Low

   No BreadcrumbList schema is present on any page, which helps AI crawlers understand site hierarchy and context.

   What to change: Add BreadcrumbList schema to all pages with breadcrumb navigation.

10. Testimonial content lacks Review schema Low

    The site has a 'What customers say about us' page with testimonials, but no Review schema is used to mark up these quotes, missing an opportunity for rich snippets.

    What to change: Add Review schema to testimonial content on the customer testimonials page.

What's working

• /en/ subdirectory pages have rich HTML content — Pages under /en/ (homepage, about, products, service) contain 300-500+ words of descriptive text, technical tables, and benefit lists, providing substantial content for AI crawlers that reach them.
• Sitemap is well-formed with 1,763 URLs — The sitemap.xml is well-formed and lists 1,763 URLs, helping crawlers discover all pages on the site.
• Robots.txt allows most AI crawlers — The robots.txt has a catch-all Allow: / and only disallows admin and contact paths, so GPTBot, Google-Extended, OAI-SearchBot, PerplexityBot, and ChatGPT-User are not blocked.
• Products subdomain contains detailed product information — The products.pfannenberg.com subdomain hosts pages with 300-800 words of product descriptions, technical specifications, and download links, providing rich data for crawlers that find it.
• Homepage leads with strong topical content on F-gas and sustainability — The /en/ homepage features 32 H2 news items about F-gas regulation, Green Series cooling units, IO-Link signaling, and sustainability, providing timely and relevant content for AI engines.
• Site links to LinkedIn, YouTube, and Xing — The site includes links to its LinkedIn (881845), YouTube, and Xing profiles, providing pathways for external signal generation.