llms.txt, schema.org and entities: what AI visibility really needs technically

A lot is being sold in the name of AI visibility right now: llms.txt files, "AI-optimised" schema, special markup for ChatGPT. If you are budgeting for GEO and AEO, prioritise the technical foundation soberly. Some measures carry measurable weight. Others are, so far, mostly hope.

What matters most for recommendations in AI search systems is crawlable, well-linked content, unambiguous entities and structured data that matches the visible text. llms.txt is a proposal from September 2024 that, according to Google's John Mueller, no AI service has officially said it uses. Google itself says AI Overviews and AI Mode require no new machine-readable files and no special schema. Prioritise entities and solid basics before new file formats.

What Google officially says about AI features

Google's documentation "AI features and your website" is unusually clear. There are no additional requirements to appear in AI Overviews or AI Mode, and no special optimisations are necessary. It states that you do not need to create new machine-readable files, AI text files or markup, and that there is no special schema.org structured data you need to add.

Instead, Google recommends the basics: allow crawling in robots.txt and via your CDN and hosting, make content discoverable through internal links, offer a good page experience, make important content available as text, align structured data with visible content and keep Merchant Center and Business Profile information up to date. For control, Google points to nosnippet, data-nosnippet, max-snippet and noindex, and to Google-Extended for AI training outside Search.

llms.txt: a proposal without confirmed use

Jeremy Howard published the llms.txt specification on September 3, 2024. The idea is a Markdown file in your site's root that gives language models a curated overview of your most important content. According to PPC Land, Ahrefs found in April 2025 that major AI services were not fetching the file in server logs. Google's John Mueller said no AI service has stated that it uses llms.txt and compared the file to the keywords meta tag.

Does that mean you should not create one? Not necessarily. The effort is small and it does no harm. But it belongs at the end of your priority list, not the beginning, and it replaces none of the fundamentals.

Entities: the underrated lever

AI systems answer questions like "Which performance marketing agency in Berlin is good?" not through individual keywords but through entities: who is this company, what does it stand for, which sources confirm it? The clearer and more consistent this information is across the web, the easier it is for a system to place you.

In practice, that means:

• Consistent core data: Name, address, services and contact details identical on your website, business profile, directories and social profiles.
• Organization markup with connections: Schema.org Organization with logo, contact details and sameAs links to your official profiles, matching visible content.
• Clear service pages: One page per service with a precise definition, instead of mixing everything on one page.
• External confirmation: Mentions in trade media, associations, directories and review platforms that repeat the same facts.

Prioritisation in four stages

1. Access: Are your pages crawled and indexed? Do your CDN, bot protection or robots.txt block relevant crawlers?
2. Content as text: Are prices, scope, processes and FAQs on the page as text, not only in PDFs or images?
3. Structure and entities: Organization, Service and FAQ markup that exactly matches visible content, plus consistent core data.
4. Extras: llms.txt and similar formats once the first three stages are in place.

Measure this not by rankings but by presence in AI answers for your most important questions and by the leads that result. How AI visibility is built and measured is covered under GEO, AEO and AI visibility, and the technical basics under SEO.

FAQ

Do I need an llms.txt for ChatGPT?

Based on current information, not necessarily. According to reports, no major AI service has officially confirmed using it. The file does no harm, but without solid basics it brings no advantage.

Does schema markup help with AI Overviews?

Google says there is no special schema required for AI Overviews. Structured data remains useful when it matches visible content and describes entities unambiguously.

What is an entity in an SEO context?

A uniquely identifiable thing, such as a company, person or product, with clear attributes and relationships. Search engines and AI systems organise information around such entities.

Sources

• Google Search Central: AI features and your website
• PPC Land: llms.txt adoption stalls as major AI platforms ignore proposed standard

Want more of the attention you get to turn into customers? Let’s spend 15 minutes on your setup, free and without a pitch deck. Book your free call