STRUCTURE THAT AI CAN READ.
- → Semantic HTML instead of div soup. AI agents parse structure, not looks.
- → Schema.org (LocalBusiness, FAQ, Service, Person, Product). Explicit entity markup for LLMs.
- → Core Web Vitals ≥ 90. Slow pages don't get cited.
THE NEW MARKET IS AGENTIC.
ChatGPT, Perplexity, Google AI Overviews and autonomous B2B buying agents are replacing classic Google search in 2026. They read pages differently than humans, and cite only what they understand. This is the technical and content guide for how your site needs to be built for that world.
SEO IS NOW AEO.
Answer Engine Optimization. Generative Engine Optimization. Agentic SEO. Different names, one fact: most of your future customers ask questions and get answers without ever opening a website. If you don't show up there, you don't exist for them.
60%
of searches end without a click in 2026 (SparkToro)
50 %
already use AI chat instead of classic search at least sometimes (Bitkom 2025)
1,151%
growth in AI traffic in 18 months (Adobe Digital Trends 2026)
44%
of LLM citations come from the first 30 % of a page
DIV SOUP IS INVISIBLE TO AI.
Classic website builders produce HTML that renders correctly for humans. Structurally it's nested, without semantic markup, without deep Schema.org, without a verifiable author. An AI agent parsing the page finds no entry point.
AIs prefer sources with traceable authorship. If that's missing, for example because a builder template runs generically for thousands of customers, the model rates the page as less citation-worthy. Even if the content is correct.
TECH → STRUCTURE → ANSWER.
EVERY TOPIC GETS ITS OWN PAGE.
- → 44% of AI citations come from the first 30 % of a page. Lead with the core answer.
- → One keyword, one URL, one schema type. Multi-page beats onepager in AI search.
- → Breadcrumbs and internal links build the topical authority LLMs look for.
AIs FIND YOU AND CITE YOU.
- → E-E-A-T signals: author schema with sameAs (LinkedIn, credentials). LLMs prefer verifiable expertise.
- → FAQPage schema per question. Direct answer snippets for AI interfaces.
- → Location and industry explicitly marked. Local AI buying agents recognize you.
WHAT THIS SITE SHIPS FOR AIs.
Every artifact below is live on medienmacher.biz. You can open it, inspect it, copy it. This runs as real production code.
01 · llms.txt
CONCENTRATED FEED FOR AI CRAWLERS.
A markdown file in the site root that summarizes brand, services, boundaries and key URLs in a machine-readable form. Generated dynamically from the WordPress theme, no manual upkeep.
→ https://medienmacher.biz/llms.txt02 · Schema.org-Graph
VERIFIABLE ENTITIES.
Every page emits JSON-LD in the head: Person, Organization, LocalBusiness, ProfessionalService, BreadcrumbList, FAQPage, Article. Stable @id references let LLMs link "Tom Zwicklhuber" across every page as the same entity.
View source on any page, search for "application/ld+json".
03 · Topic cluster
ONE PAGE PER TOPIC.
Packages, Headless, Rescue, AI Audit, Agentic SEO, About, FAQ, Certificates, Work. Each on its own URL, its own schema type and a canonical topic anchor. That's the structure LLMs weight authority against.
04 · Answer chunks
FAQs AS DIRECT QUOTES.
Over 30 FAQs with FAQPage schema, grouped into categories (process, pricing, tech, support, KMU.DIGITAL funding, vertical-specific). Generative engines pull cite-ready snippets from here. Answers are short, unambiguous, and explicit about what I do NOT do.
→ /faq/FIND OUT IN ONE REPORT.
The AI audit tests your site against exactly these requirements: LLM cite check, Schema.org, BFSG/EAA accessibility, machine-readability. flat €490.
VIEW AI AUDIT →