Skip to content
Book a call

3 min readSEO

Getting cited by ChatGPT and Perplexity: what llms.txt and structured data really do

What makes AI search quote a site: being crawlable and indexed, plain answers, structured data and mentions elsewhere. And what llms.txt does and doesn't do.

AI search tools quote pages they can fetch, find in a search index and understand quickly. So the work is mostly good SEO done properly: let their crawlers in, render your pages on the server, be indexed by Google and Bing, answer each question in plain sentences near the top, and describe yourself with structured data. llms.txt is a cheap extra, not a shortcut. Mentions on other sites matter as much as anything on your own.

How AI search finds a page

When ChatGPT, Perplexity or Google's AI Overviews answer with sources, they search an index, fetch the most relevant pages and quote from them. ChatGPT's search relies on outside search providers, Bing among them, and on its own crawler; Google's AI features use Google's index. A page that isn't indexed, or that returns an empty shell without JavaScript, can't be quoted.

What to check, in order

  1. Crawlers are allowedYour robots.txt shouldn't block OAI-SearchBot, PerplexityBot or Googlebot. Blocking GPTBot (training) is a separate choice from blocking search.
  2. Pages render on the serverFetch your page with JavaScript off. If the text isn't there, most AI crawlers don't see it either. This is the most common problem with apps built in Lovable, Bolt and v0.
  3. Indexed in Google and BingSubmit a sitemap to both Search Console and Bing Webmaster Tools, and use IndexNow so Bing hears about new pages at once.
  4. Answer firstPut the direct answer to the page's question in the first paragraph, then the detail. AI tools quote sentences that stand on their own.
  5. Structured dataOrganization or Person with sameAs links to your profiles, plus Article, Service and BreadcrumbList. It tells engines who wrote the page and what it is about.
  6. Mentions elsewhereDirectories, community answers, guest posts and profiles that link to you. AI answers often cite forums and lists, and those in turn point to the sites they trust.

What llms.txt is, and isn't

llms.txt is a proposed convention: a Markdown file at your site's root that summarises what the site is and links to its key pages, written for language models. It's quick to add and useful to developer tools and agents that read it. No major AI search engine has said it uses llms.txt for ranking or citations, so treat it as a helpful summary, not a lever.

This site generates its llms.txt from the same content as the pages, so it never goes stale, and the on-site assistant answers from it too.

How to measure it

  • Ask ChatGPT, Perplexity and Google the questions your buyers ask, with search on, and note who gets cited. Repeat monthly.
  • Watch referrals from chatgpt.com and perplexity.ai in your analytics.
  • Bing Webmaster Tools now reports how often your pages appear in Copilot answers.
My AI search audit checks all of this on your site and ranks the fixes; on Next.js sites I make them too. See AI search and SEO.

Questions people ask

What is generative engine optimisation (GEO)?
Making your site easy for AI search tools to fetch, understand and quote. Most of it overlaps with technical SEO: server-rendered pages, indexing, clear answers and structured data, plus being mentioned on other sites.
Does llms.txt help me rank in ChatGPT?
There's no public evidence that ChatGPT or Perplexity use llms.txt for ranking or citations. It's cheap to add and helps some tools, but crawlability, indexing and clear answers matter far more.
Why doesn't ChatGPT mention my website?
Usually because the pages aren't indexed, render empty without JavaScript, block AI crawlers, or don't answer the question directly. Few mentions on other sites is the other common reason.
Should I block GPTBot?
That's a choice about training, not search. Blocking GPTBot stops OpenAI using your pages to train models; allowing OAI-SearchBot is what lets ChatGPT search show and cite them.

Reply within a day

Got an app, or an idea?

Send me the link, or a few lines about what you want to build. I'll reply within a day with honest next steps, even if I'm not the right fit.

Free 30-minute call · No commitment · Reply within a day

Pick a time on my Cal.com page. Calls are on Google Meet. Or email me.

  1. Step 1

    You send a link or an idea

    A few lines is enough.

  2. Step 2

    I reply within a day

    With questions or a first take.

  3. Step 3

    A free 30-minute call

    Then a written quote.