Guide

How to optimise your site for ChatGPT

When someone asks ChatGPT for a recommendation in your space, it can only mention sites it can actually reach and read. Most sites fail at the basics — not because the fixes are hard, but because nobody has checked. This guide covers the eight things that matter most, in priority order. Every one of them is something you can verify yourself in a few minutes.

01Let the AI crawlers in

ChatGPT, Claude and Perplexity all crawl the web with their own bots — GPTBot, ClaudeBot and PerplexityBot. Many sites block them without realising, often through a copied robots.txt rule or a hosting default.

Open yourdomain.co.uk/robots.txt and look for any Disallow: / rule that applies to these bots. If you want AI tools to cite your content, they need access:

User-agent: GPTBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: PerplexityBot
Allow: /

02Publish an llms.txt file

An llms.txt file is a plain-language summary of your site written for language models — who you are, what you do, and which pages matter most. It sits at yourdomain.co.uk/llms.txt and gives AI tools a clean, unambiguous starting point instead of leaving them to guess from your HTML.

Keep it short: a one-line description, then a list of your most important pages with a sentence about each. Markdown is fine. This is the single highest-impact change most sites are missing.

03Add structured data

Structured data (JSON-LD) tells machines exactly what a page is about — your organisation, your services, your articles — in a format they can parse without interpretation. AI search tools lean on it heavily when deciding what a site is and whether it's a credible answer.

At minimum, add an Organization or WebSite schema to your homepage, and Article schema to any content pages. It lives in a script tag in your page head:

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Your Company",
  "url": "https://yourdomain.co.uk"
}
</script>

04Keep a sitemap — and reference it

A sitemap at /sitemap.xml helps every crawler — traditional and AI — find all of your pages, not just the ones they stumble across. The step most sites miss: reference it from robots.txt so crawlers don't have to guess where it lives.

Sitemap: https://yourdomain.co.uk/sitemap.xml

05Write a real meta description

Your meta description is often the text AI tools quote when summarising what you do. Aim for 50–160 characters that plainly state what the page offers. If it's missing, or stuffed with keywords, AI summaries of your site will be vague — or wrong.

06Get your title tags right

Every page needs a unique title of 10–70 characters that names what the page is. Titles are one of the strongest relevance signals for both classic search and AI answers. "Home" tells a machine nothing; "Ceramic dinnerware handmade in Bristol — Studio Name" tells it everything.

07Complete your Open Graph tags

Open Graph tags (og:title, og:description, og:image) control how your pages appear when shared — and they give AI tools a second, consistent source of truth about each page. All three should be present on every page that matters.

08Use a clean heading structure

One h1 per page, then h2s and h3s in order without skipping levels. Language models use headings to understand the shape of your content — a page with three h1s and jumps from h2 to h5 is harder to summarise accurately.

Check your site in about ten seconds

The free scanner runs all eight of these checks against your URL and gives you a prioritised fix list. No account needed.

Scan your site free