The GetFreeAEO starter kit

Let the AI crawlers in

AI models learn about your site from crawlers with names like GPTBot and ClaudeBot. Many sites block them without knowing it, either in robots.txt or at the hosting edge. If the crawlers cannot get in, the models are working from memory and guesswork.

Add explicit allow rules for the major AI crawlers to your robots.txt, at the root of your site:

User-agent: GPTBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Google-Extended
Allow: /

User-agent: OAI-SearchBot
Allow: /

If your hosting provider or CDN blocks bots at the edge, the robots.txt rules alone will not be enough. That is a ticket to your hosting vendor, not a file edit.

Add an llms.txt file

Models check for an llms.txt file at your site root the way browsers check for favicons. It is a plain markdown file that tells AI, in your own words, what your company does and where the important pages are. Without it, the model pieces your story together from scattered pages.

Create a file called llms.txt at the root of your site (so it loads at yoursite.com/llms.txt). Keep it short and factual:

# Example Company

> Example Company sells handmade leather goods from Portland, Oregon.

## Products

- [Leather tote bags](https://example.com/products/totes)
- [Card wallets](https://example.com/products/wallets)

## Company

- [Our story](https://example.com/about)
- [Shipping and returns](https://example.com/shipping)

Larger sites can add an llms-full.txt with the complete version. One page of accurate facts beats ten pages of marketing copy.

Tell the models who you are

When a shopper asks an AI about your category, the model needs a canonical answer for who you are, where you operate, and what you sell. Organization schema gives it one, in a format it trusts over page copy.

Add this JSON-LD block to your homepage, with your real details filled in. One detail matters more than people expect: use the same business name, address, and phone number everywhere, your site, your profiles, your directories. Mismatches make models hedge.

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Example Company",
  "url": "https://example.com",
  "logo": "https://example.com/logo.png",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "123 Main St",
    "addressLocality": "Portland",
    "addressRegion": "OR",
    "postalCode": "97201",
    "addressCountry": "US"
  }
}

Answer the questions buyers actually ask

Shoppers ask AI direct questions: how much, how fast is shipping, what is the return policy, which product fits my need. If your site never states the answers plainly, the model answers from a competitor that does.

Write a short FAQ on each key page, answering in one or two sentences each, then mark it up with FAQPage schema so models can lift the answers verbatim:

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [{
    "@type": "Question",
    "name": "How long does shipping take?",
    "acceptedAnswer": {
      "@type": "Answer",
      "text": "Orders ship in 1 to 2 business days."
    }
  }]
}

Do not invent numbers. If a claim has no published source on your site, like a satisfaction rate or a delivery time, either publish the methodology or take the claim down. Models repeat what you publish, including the parts you cannot defend.

Write your titles for the question

Your title tag is still the single strongest signal of what a page is about, for search engines and for AI. Most titles are written for humans browsing tabs. Write them for the question being asked.

A product or service page title should state what it is and who it is for, in plain words: "Leather Tote Bags, Handmade in Portland | Example Company" beats "Collection No. 4". Keep meta descriptions to one factual sentence about the page. No slogans, no keyword stuffing.