Getting cited by ChatGPT

To get cited by ChatGPT, make your answer easy to retrieve, easy to lift and easy to trust. ChatGPT does not run JavaScript, so your content and schema must be in the server-rendered HTML. Allow its crawlers, lead with a direct answer, cite your own sources, keep pages fresh, and build a presence on third-party sites ChatGPT already reads.

How does ChatGPT decide what to cite?

ChatGPT decides what to cite the same way the other engines do: it retrieves candidate pages, extracts the passage that best answers the question, and prefers sources it can trust enough to name. Two things make ChatGPT distinct. It does not run JavaScript, so it only sees your raw HTML, and it synthesises the wider web, so third-party mentions of you matter alongside your own pages. Everything below follows from those two facts. For the general mechanism, see how-ai-engines-choose-what-to-cite.

Make sure ChatGPT can read your page

The first requirement is that ChatGPT can reach and read the page. It reads raw HTML and skips JavaScript, so serve your content and schema server-side.

Give ChatGPT a liftable answer

Once ChatGPT can read the page, give it a passage it can lift. Lead with a direct 40 to 70 word answer to the question the page addresses, in one self-contained paragraph, before any preamble. Use question-shaped headings that match how the query is phrased, open each section with its own direct answer, and put comparative facts in tables or lists. Engines quote the first clean, self-contained passage they find, so an answer-first page is far easier to cite than one that buries the point. See answer-first-content.

Give ChatGPT reasons to trust you

ChatGPT prefers sources it can stand behind, so give it evidence. Cite your own sources with outbound links, quote statistics and attribute them, and show a named author with real credentials. This lines up with the Princeton GEO study (KDD 2024), which found that citing sources raised visibility in generative answers by roughly 40 percent, quoting statistics by about 37 percent, and adding quotations by about 30 percent, while keyword stuffing lowered it by around 10 percent. Sourced, well-attributed content is more citable than confident but unsupported prose.

Keep your pages fresh

Freshness helps, because AI crawlers favour recent content. The large majority of AI-bot crawl activity is on recent content (Seer, 2025), so a page that is updated and dated is more likely to be retrieved and cited than one that looks stale. Keep important pages current, show a visible last-updated date, and refresh the facts and statistics rather than leaving them to age. Freshness is one of the Trusted-pillar checks Rubric scores, flagging pages not updated within twelve months.

Build third-party presence

ChatGPT synthesises the wider web, so being described on trusted third-party sites raises your chance of a citation. Coverage, mentions and consistent descriptions of your organisation on sites ChatGPT already reads all feed the picture it builds of you. This is off-page work, and it is worth being honest about a limit: Rubric estimates on-page citability and does not measure off-page authority. Use Rubric to get your own pages citable, then pursue third-party presence separately.

How Rubric helps you get cited by ChatGPT

Rubric helps by scoring your pages the way the no-JS engines, including ChatGPT, actually read them. It weights ChatGPT among the six scored engines, checks that your schema is readable without JavaScript, that the AI bots are allowed, that your answer is liftable, and that your evidence is sourced and fresh. You can lint a draft before you publish it, or audit a live site and work the phased Action Plan. To learn how to read the resulting score, see reading-your-citability-score, and for how ChatGPT compares with the other engines, see how-the-six-ai-engines-differ.