weisenburger.de — AI Site Grade

Analysis

The sitemap sub-sitemaps all redirect to the homepage — that's a major structural finding. The cold LLM knowledge also has a notable inaccuracy (Stuttgart 21 involvement). Let me write the audit.

weisenburger.de: AI-Visibility Audit

The sitemap index points to five sub-sitemaps (pages, news, references, blogPosts, realEstates), but every one of those sub-sitemap URLs redirects to the homepage with a 200 and zero URLs listed — effectively a broken sitemap architecture that leaves all deeper content invisible to crawlers relying on sitemap discovery.

Crawler Access

All major AI bots (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, anthropic-ai, Bytespider, Applebot-Extended) receive identical 200 responses as a browser, with full HTML content at ~94KB. No UA-based blocking exists. The site runs on Apache (no CDN/WAF layer) with a strong Content-Security-Policy but no Strict-Transport-Security and no X-Frame-Options. The robots.txt is minimal — a single User-Agent: * rule disallowing /typo3/ and /musterhaus_data/ — with no AI-bot-specific directives whatsoever. The /llms.txt returns a 404 (TYPO3 error page). No JS-rendering risk: plain GET returns full visible text (~900 words on the homepage).

Cold-Knowledge Gap

The LLM prior describes Weisenburger as a Karlsruhe-based construction firm involved in the "controversial Stuttgart 21 railway project" and facing "criticism over alleged construction defects and delays" in 2023. The actual site makes zero mention of Stuttgart 21 or any legal scrutiny. The site positions itself as "the leading medium-sized general contractor in southern and western Germany," "cost leader in terraced house construction," and a 70-year-old family-run group with 650+ employees and 500M EUR annual output. The cold knowledge is stale, inaccurate, and missing the entire group structure (bau, projekt, invest, technischer service, sanierungsprofi).

Schema Posture

Every page carries only BreadcrumbList and WebSite schema types. No Organization, LocalBusiness, Product, FAQPage, or RealEstateListing schema exists anywhere — despite the site offering detailed property listings with prices, square meters, room counts, and status indicators on /immobilienangebot. The homepage has no Organization schema with logo, social profiles, or contact data. The breadcrumb schema is present but structurally shallow (always 1-2 levels).

Content & External Signals

The homepage is a news-heavy landing page with 30+ dated press entries (from 2024 through May 2026), pushing the core value proposition below the fold. The site has a bilingual German/English structure with full content parity. External signals are thin: no significant press coverage, Reddit threads, or review sites surfaced in searches. The site links to kununu, LinkedIn, Xing, Facebook, Instagram, and YouTube — but these are footer links, not structured sameAs schema references. The blog lives at mein-weisenburger-haus.de (a separate domain), and the renovation arm at sanierungsprofi.de — both off-domain and not cross-linked in any structured way.

Findings

1. All sub-sitemaps redirect to homepage, listing zero URLs High

   The sitemap index points to five sub-sitemaps (pages, news, references, blogPosts, realEstates), but every sub-sitemap URL redirects to the homepage with a 200 status and zero URLs listed. This leaves all deeper content invisible to crawlers relying on sitemap discovery.

   What to change: Fix the sub-sitemap generation so each sub-sitemap returns a valid XML listing of its respective URLs instead of redirecting to the homepage.

2. Cold LLM knowledge contains stale inaccuracies about Stuttgart 21 involvement High

   The LLM prior describes Weisenburger as involved in the controversial Stuttgart 21 railway project and facing criticism over construction defects and delays in 2023. The actual site makes zero mention of Stuttgart 21 or any legal scrutiny, and the cold knowledge is missing the entire group structure.

   What to change: Publish a clear, factual company overview page with group structure and project history to correct the LLM's stale knowledge.

3. No Organization or LocalBusiness schema on any page High

   Every page carries only BreadcrumbList and WebSite schema types. No Organization, LocalBusiness, Product, FAQPage, or RealEstateListing schema exists anywhere, despite the site offering detailed property listings with prices, square meters, room counts, and status indicators on /immobilienangebot.

   What to change: Add Organization schema with logo, social profiles, and contact data to the homepage, and add RealEstateListing schema to property detail pages.

4. /llms.txt returns 404 error page Medium

   The /llms.txt endpoint returns a 404 (TYPO3 error page), meaning there is no LLM-specific content file to guide AI crawlers to key pages or provide a concise summary.

   What to change: Create an /llms.txt file that lists key pages (e.g., company overview, services, projects) and provides a brief summary of the site.

5. Robots.txt has no AI-bot-specific directives Medium

   The robots.txt contains a single User-Agent: * rule disallowing /typo3/ and /musterhaus_data/, with no AI-bot-specific directives (e.g., GPTBot, ClaudeBot). While all bots currently get full access, the lack of explicit allowances leaves room for future default blocking.

   What to change: Add explicit Allow directives for AI bots (e.g., GPTBot, ClaudeBot) to ensure continued access as default policies evolve.

6. Missing Strict-Transport-Security and X-Frame-Options headers Low

   The site runs on Apache with a strong Content-Security-Policy but lacks Strict-Transport-Security and X-Frame-Options headers, which are not critical for AI visibility but indicate suboptimal security posture.

   What to change: Add Strict-Transport-Security and X-Frame-Options headers to improve security posture.

7. Homepage is news-heavy, pushing core value proposition below the fold Medium

   The homepage features 30+ dated press entries (from 2024 through May 2026), which pushes the core value proposition (leading medium-sized general contractor, cost leader in terraced house construction) below the fold, potentially diluting the key message for AI crawlers.

   What to change: Restructure the homepage to prominently feature the core value proposition and key services above the news section.

8. Social media links in footer lack structured sameAs schema Low

   The site links to kununu, LinkedIn, Xing, Facebook, Instagram, and YouTube in the footer, but these are not referenced via structured sameAs schema, missing an opportunity to strengthen entity recognition.

   What to change: Add sameAs properties to the Organization schema pointing to all social media profiles.

9. Blog and renovation arm on separate domains without structured cross-linking Medium

   The blog lives at mein-weisenburger-haus.de and the renovation arm at sanierungsprofi.de, both off-domain and not cross-linked in any structured way, fragmenting the site's authority and making it harder for AI to connect related content.

   What to change: Consider consolidating key content under the main domain or adding structured cross-links (e.g., sitemap entries, schema references) between domains.

10. Thin external signals with no significant press or review coverage Medium

    Searches for Weisenburger in news, reviews, and criticism returned zero results, indicating low external visibility and limited third-party validation that AI models might use to assess authority.

    What to change: Encourage customer reviews on platforms like Google Maps and kununu, and pursue press coverage to build external signals.

What's working

• All major AI bots receive full HTML content without blocking — All tested AI bots (GPTBot, ClaudeBot, PerplexityBot, etc.) receive identical 200 responses with full HTML content (~94KB), and no UA-based blocking exists. The robots.txt does not disallow any AI bot.
• Bilingual German/English structure with full content parity — The site has a bilingual German/English structure with full content parity, making it accessible to a wider audience and AI models that may prefer English.
• BreadcrumbList schema present on all pages — Every page carries BreadcrumbList schema, which helps AI understand site structure and navigation paths.
• Plain HTML content without JavaScript rendering requirements — All pages return full visible text in the initial HTML response (~900 words on homepage), so AI crawlers do not need to execute JavaScript to access content.
• Strong Content-Security-Policy header in place — The site has a strong Content-Security-Policy header, which helps prevent XSS attacks and ensures content integrity.