llms.txt, Schema and Citability: the Technical AEO Checklist

Content earns citations, but the technical layer decides whether engines can even see that content. This is the part of AEO you can finish in a week and never rewrite from scratch again.

llms.txt: a table of contents for AI

llms.txt is a plain-markdown file at your site root that tells language models what your site is and where the important pages are. Think of it as robots.txt's constructive sibling: instead of saying "keep out", it says "here's what matters".

A minimal useful llms.txt:

# YourBrand

> One-paragraph factual description of what the product
> or company does, for whom, and what makes it distinct.

## Key pages

- [Product overview](https://example.com/product): what it does
- [Pricing](https://example.com/pricing): plans and limits
- [Docs](https://example.com/docs): integration guide
- [Blog](https://example.com/blog): guides and research

Keep it factual and short — it is a summary for machines, not a landing page. Update it when your site structure changes. SERP Checker publishes its own at /llms.txt, and our Tech-SEO module checks clients' sites for one.

The schema kit that pays for itself

Four JSON-LD types cover most sites:

• Organization — home page: name, logo, URL, social profiles. Anchors your entity.
• Article — every post: headline, description, dates, author. Engines strongly prefer dated, attributed content.
• FAQPage — any page with Q&A. The question/answer pairs map one-to-one onto how answer engines compose responses.
• Product — commerce pages: name, price, availability, ratings.

Two rules: JSON-LD in the head (not microdata scattered through markup), and markup must mirror visible content — engines validate, and fabricated schema erodes trust instead of building it.

Crawler access: the boring check that decides everything

• robots.txt returns 200 and does not block GPTBot, ClaudeBot, PerplexityBot, Google-Extended.
• Key content renders server-side. Most AI crawlers execute little or no JavaScript: if your text arrives via client-side rendering, for them the page is empty.
• sitemap.xml is current, referenced from robots.txt, free of redirects and 404s.
• TTFB under ~500 ms. AI crawlers have tighter budgets than Googlebot; slow origins get partially indexed.

The checklist

1. llms.txt published at site root, linked pages return 200.
2. robots.txt allows the four major AI crawlers.
3. Organization schema on the home page; Article on posts; FAQPage on Q&A pages.
4. Server-side rendering (or prerendering) for all content pages.
5. sitemap.xml current and clean.
6. TTFB < 500 ms on key pages.
7. Every key page passes the ten-second answer test — a stranger can lift a complete answer fast. More on that in the citation playbook.

Want the automated version? The free SERP Checker audit runs the technical half of this list against any URL in about a minute — schema, robots, sitemap, speed, structure — and scores it 0–100.

FAQ

Is llms.txt an official standard?

It is a community convention (llmstxt.org), not an IETF standard — but adoption among AI tooling is growing, it costs nothing, and it has no downside. Early adopters get disproportionate clarity in how models describe them.

Will schema markup alone get me cited?

No. Schema makes your content unambiguous to machines; it does not make it worth citing. Pair it with answer-shaped content and a verifiable brand entity — schema is the amplifier, not the signal.

How do I know if AI crawlers actually visit my site?

Check server logs for GPTBot, ClaudeBot, PerplexityBot and Google-Extended user agents. No visits in weeks usually means a robots.txt block, a firewall rule, or an origin too slow to bother with.