October 20, 2026

Structured Data and llms.txt: The Technical Basics for AI Search

Structured Data and llms.txt: The Technical Basics for AI Search

The technical basics that help AI search understand your company are structured data on your website, clean and readable HTML, and crawler access that is not accidentally blocked. llms.txt is a newer, optional file that some sites add, but it is a community proposal, not an established standard, and the major search providers have not confirmed that they rely on it. Fix the fundamentals first, then decide about llms.txt.

This is the technical layer of what we call generative engine optimization. If the concept is new to you, start with what GEO is and how it works.

Structured data: telling machines what your page means

Structured data is a small block of code, usually in JSON-LD format, that describes your content using the shared vocabulary of schema.org. A human sees "FADETOBLACK, Munich, video production". Structured data states it explicitly: this is an organization, this is its address, these are its services, this is the founder. Google documents structured data as a way to help it understand page content, and the same clarity helps any system that reads your site.

The types that matter most for a service business:

  • Organization or LocalBusiness: name, logo, address, contact, and links to your official profiles via the sameAs property. This connects your website to your LinkedIn, Instagram and YouTube presence.
  • Service: what you offer, for whom, and where.
  • Person: founders and experts with their role, so authorship and expertise are attributable.
  • Article or BlogPosting: author, publication date and topic for every article.
  • FAQPage: questions and answers in a machine readable form. Google restricted FAQ rich results in search to a small group of authoritative sites in 2023, but the markup still describes your content clearly.

Rule of thumb: structured data must match what is visible on the page. Marking up claims that are not on the page is against Google's guidelines and undermines trust.

The basics that are often forgotten

  • Crawler access: check your robots.txt. AI providers use their own crawlers, for example OpenAI's GPTBot and OAI-SearchBot. Blocking them by default, sometimes via a security setting you never chose, makes you invisible to those systems.
  • Text in HTML, not only in images or scripts: key statements about your services should exist as real text in the page, not only inside a video, a PDF or content that only appears after scripts run.
  • Clear page structure: one topic per page, meaningful headings, a direct answer in the first paragraph.
  • Consistent facts: the same company name, address and service description on your site, Google Business Profile and social profiles.

What llms.txt is, and what it is not

llms.txt was proposed in 2024 by Jeremy Howard of Answer.AI. It is a Markdown file at the root of your domain that summarises your site and links to the most important pages in a format that language models can read easily. It is simple to create and does no harm.

But be realistic. It is not part of any official standard, and Google representatives have publicly said Google does not use it for search. Treat it as a low effort extra, useful for documentation heavy sites and developer tools, not as the lever that gets you into ChatGPT answers. Your structured data, your content and your presence on sources AI systems trust matter much more. You can test your current status with our guide Does your business appear in ChatGPT?

A short checklist

  • Organization or LocalBusiness markup with sameAs links on the homepage.
  • Service markup on each service page, Article markup on each blog post.
  • robots.txt reviewed for AI crawlers.
  • Core statements as visible text with clear headings.
  • Optional: an llms.txt that lists your key pages.

FADETOBLACK implements these technical foundations as part of our AI search visibility work in Munich, together with the content that AI systems actually cite. If you want to know where your site stands, we will review it with you.

Book a free initial consultation