AI search
How to get cited by ChatGPT, Perplexity, and Gemini
· Configuration notes, not a newsroom.
Answer engines cite pages they can fetch and ground. Block GPTBot and you are invisible. Ship llms.txt, entity schema, and dated articles — then rescan.
They cannot cite what they are not allowed to fetch
ChatGPT browsing uses GPTBot and ChatGPT-User. Perplexity uses PerplexityBot. Claude uses ClaudeBot. Gemini grounding respects Google-Extended as an opt-out. If robots.txt Disallow: / those names, no amount of 'AI content' on the blog will show up as a live citation.
SiteRune parses the file. We do not impersonate the bots. The finding is aeo-bots-blocked when a named crawler is locked out.
Give them a map and an entity
llms.txt should list 5–15 durable URLs with one-line facts. Organization JSON-LD needs a name, url, and preferably a logo. FAQPage or TechArticle on a real URL beats a homepage slogan. Dates on articles (datePublished in JSON-LD and a visible time element) stop models treating you as undated folklore.
Pages that look like answers
Put some H2/H3 as actual questions. Answer in the first 60 words, then the nuance. That is how extractors build Q→A pairs. Marketing headlines ('Unleash your stack') do not parse. SiteRune's aeo-questions finding is the cheap version of that check.
Do not buy fake citations
We will not recommend paying for Reddit or Quora placement as a SiteRune fix. If you want a citation, make the origin fetchable and factual. Run a scan, ship the playbook, rescan. The delta is the product.
Run it on a live URL
The brief will name your CMS and attach the files. Three guest scans, no card.