elten.com — AI Site Grade

Analysis

ELTEN.com: AI crawlers get full access to a site that delivers almost no readable content

All 11 AI crawlers tested (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Bytespider, Applebot-Extended, anthropic-ai, Perplexity-User, and a plain Browser) receive a 200 status with identical byte payload (387,750 bytes) from the homepage — no UA-based blocking, no WAF challenge, no Cloudflare wall. Yet the visible text extracted from the homepage is only ~90 words. The page is a JavaScript-heavy WordPress shell (Enfold theme, Yoast SEO Premium) where the main content — product names, descriptions, navigation — is rendered client-side or loaded via lazy images. AI crawlers get the HTML skeleton but almost no substantive text about products, specifications, or brand value.

Crawler Access

The robots.txt at https://elten.com/robots.txt contains a single User-agent: * rule with 40+ Disallow paths (dealer portals, old uploads, tag pages, /aktuelles/, /stage/, /wp-admin/). No AI-specific bot directives exist — no GPTBot, ClaudeBot, Google-Extended, or PerplexityBot rules. The llms.txt at https://elten.com/llms.txt returns a 404 (serving the full WordPress 404 page, 315KB). The site runs on a plain Apache server at 85.13.133.50 (Intersolute hosting, Germany) with no CDN, no WAF, no security headers (no HSTS, no CSP, no X-Frame-Options), and a Cache-Control: max-age=3 header — effectively uncached.

Content and Schema Posture

Every page examined — homepage, /produkte/, /produktgruppen/fire/, /produktgruppen/wellmaxx/, blog posts — returns ~90 words of visible text from a plain GET. The body is dominated by the footer navigation (product group names, service links) and an accessibility widget. The actual product descriptions, technology explanations, and brand storytelling are embedded in images or loaded via JavaScript. JSON-LD schema is present via Yoast but limited to WebPage, BreadcrumbList, WebSite, and ImageObject types. No Product schema, no Organization schema (with logo, address, founding date), no FAQPage schema despite FAQ content being linked. The English homepage at /en/ has no meta description. Blog posts have author schema (Christopher Neumann) and dates, but the text content is again minimal.

Cold-Knowledge Gap

The LLM knows ELTEN as a German safety footwear manufacturer founded in 1918, based in Tönisvorst, manufacturing entirely in Germany, with specific model names (RENZO, MILAN) and sole technology (ELTEN Sportive). The actual site says the company is based in Uedem (Ostwall 7-13, 47589 Uedem), not Tönisvorst — a location mismatch in the model's training data. The site also prominently features the LOWA WORK COLLECTION partnership, mentions Infinergy by BASF, BOA fit system, GORE-TEX CROSSTECH, and product lines like WELLMAXX, BIOMEX, and FIRE & RESCUE — none of which surfaced in the cold knowledge. The model's "Made in Germany" emphasis is correct but the site itself never explicitly states this on any fetched page.

External Signals

External search results for ELTEN returned zero results from DuckDuckGo for queries in both English and German — an anomaly that suggests the site has low external citation density or the search engine is not indexing it well. No Reddit threads, no review sites, no trade press articles appeared in search. The site links to Facebook, Instagram, YouTube, and LinkedIn profiles, plus an external store at elten-store.de and a dealer portal at dealerportal.elten.com. The Wayback Machine shows a snapshot from March 2026, confirming the site is live and updated. The DNS TXT records include Google and Apple domain verification tokens, indicating the brand has claimed its web presence with search engines.

Structural Surprise

The sitemap index at https://elten.com/sitemap_index.xml contains 30 sub-sitemaps covering products, portfolio entries, categories, and filter taxonomies (color, size, closure type, protection class, gender) — a rich information architecture that should be an AI goldmine. But the actual pages deliver almost no parseable text. The site has a massive content extraction problem: AI crawlers can access everything but understand almost nothing because the substantive content is image-based or JS-rendered. The /aktuelles/ (news) path is explicitly disallowed in robots.txt, blocking crawlers from what could be timely brand content.

Findings

1. Pages render as JavaScript shells with ~90 words of visible text High

   Every page examined returns a 200 status with ~90 words of extractable text. The main content—product names, descriptions, specifications—is embedded in images or loaded client-side, making it invisible to AI crawlers.

   What to change: Implement server-side rendering or static HTML for key content areas, especially product descriptions and specifications. Ensure that text is not solely delivered via JavaScript or images.

2. No Product or Organization schema on product pages High

   JSON-LD schema is limited to WebPage, BreadcrumbList, WebSite, and ImageObject. Product pages lack Product schema, and the site lacks Organization schema with logo, address, and founding date.

   What to change: Add Product schema to all product pages with name, description, SKU, brand, and offers. Add Organization schema to the homepage with name, logo, address, and founding date.

3. llms.txt returns 404 Medium

   The llms.txt file at https://elten.com/llms.txt returns a 404 status, serving the full WordPress 404 page (315KB). This is a missed opportunity to provide AI crawlers with a structured summary of the site.

   What to change: Create an llms.txt file at the root that lists key pages, product categories, and a brief description of the site for AI crawlers.

4. No AI-specific bot directives in robots.txt Low

   The robots.txt file contains only a generic User-agent: * rule with Disallow paths. No directives for GPTBot, ClaudeBot, Google-Extended, or other AI crawlers exist, leaving them unrestricted but also unguided.

   What to change: Add specific rules for AI crawlers to allow access to key content while disallowing low-value paths like /wp-admin/ and /aktuelles/.

5. Aktuelles (news) path disallowed in robots.txt Medium

   The /aktuelles/ path is explicitly disallowed in robots.txt, blocking crawlers from accessing timely brand news and updates that could improve AI visibility.

   What to change: Remove the Disallow rule for /aktuelles/ or allow specific AI crawlers to access it.

6. Zero external search results for ELTEN across multiple queries High

   DuckDuckGo searches for ELTEN safety shoes, brand reviews, and 'Made in Germany' returned zero results. This indicates low external citation density and poor indexing, limiting AI knowledge of the brand.

   What to change: Implement a link-building strategy to earn mentions on review sites, trade publications, and social media. Ensure the site is properly indexed by search engines.

7. LLM training data lists wrong headquarters location Medium

   The LLM knows ELTEN as based in Tönisvorst, but the site states the company is based in Uedem (Ostwall 7-13, 47589 Uedem). This mismatch can cause AI to fabricate incorrect information.

   What to change: Add Organization schema with the correct address to the contact page and homepage to help AI models update their knowledge.

8. English homepage lacks meta description Medium

   The English homepage at /en/ has no meta description, reducing its visibility in search engine results and limiting the context AI crawlers can extract.

   What to change: Add a unique meta description to the English homepage that summarizes the brand and its offerings.

9. No FAQPage schema despite FAQ content Medium

   The site links to FAQ content but does not use FAQPage schema, missing an opportunity to appear in rich results and provide structured answers to AI crawlers.

   What to change: Add FAQPage schema to any page containing frequently asked questions.

10. Cache-Control max-age=3 and no CDN Low

    The site sets Cache-Control: max-age=3 seconds and runs on a plain Apache server without a CDN, resulting in poor performance and potential crawl budget waste.

    What to change: Increase cache duration for static assets and consider using a CDN to improve load times and crawl efficiency.

What's working

• All 11 AI crawlers receive 200 status with no blocking — No UA-based blocking, WAF, or Cloudflare challenges are applied to AI crawlers. All tested bots receive the same 200 response as a regular browser.
• Sitemap index with 30 sub-sitemaps covering products and categories — The sitemap index contains 30 sub-sitemaps for products, portfolio entries, categories, and filter taxonomies, providing a comprehensive map of the site's content.
• Google and Apple domain verification tokens in DNS — DNS TXT records include verification tokens for Google and Apple, indicating the brand has claimed its web presence with search engines.
• Yoast SEO Premium provides basic schema markup — The site uses Yoast SEO Premium, which injects JSON-LD schema for WebPage, BreadcrumbList, WebSite, and ImageObject on most pages.
• Blog posts include author schema and publication dates — Blog posts have author schema (Christopher Neumann) and dates, providing structured metadata that AI crawlers can use.
• Links to Facebook, Instagram, YouTube, LinkedIn, and external store — The site provides links to its social media profiles and an external store (elten-store.de), offering additional channels for brand visibility.