The one file that tells ChatGPT how to read your site

Last week we talked about the workflow for deciding what to fix. This week, one concrete fix that almost no small site has done yet — which is exactly why it's worth doing.

You already know robots.txt: the file that tells Google's crawler what it can and can't look at. There's now a younger sibling aimed at AI tools — llms.txt.

Why this suddenly matters

AI referral traffic — people arriving from ChatGPT, Perplexity, Gemini — grew over 350% year on year, and ChatGPT alone now drives roughly a fifth of search-related traffic worldwide. Those tools send crawlers to read your site before they answer. If your content is hard for them to parse, you simply don't get quoted.

llms.txt is a plain-text file at the root of your domain that points AI crawlers to your most important, cleanest content — like a menu handed to a guest instead of making them wander the kitchen.

The fix (about 15 minutes)

1. Create a file called llms.txt.
2. At the top, add your site name and a one-line description of what you do.
3. List your highest-value pages as markdown links — core services, your best guides, key product pages — each with a short note on what it covers.
4. Upload it to your site root so it's reachable at yourdomain.com/llms.txt. On WordPress you can do this via your host's file manager or a small plugin.
5. Keep it short and current. This is a curated menu, not a sitemap dump.

Why it works: you're removing guesswork for the crawler. Clear structure and clean pointers make it far more likely your wording — not a competitor's — ends up in the AI answer a buyer reads.

Want to know how you read today?

Before you write the file, it helps to see how machines currently interpret your site. The AI Visibility Score runs a five-pillar check in a few minutes. And if AI search is becoming a real channel for you, the AI Visibility Sprint is a focused piece of work to get the whole setup right — structured data, content clarity, and crawler files included.

---

Every issue lives in the archive, organised by topic. See you next week.