How to Get Your Business Cited by ChatGPT
ChatGPT answers from sources it trusts, and that set changed in late 2025. Here is what it draws on now and how a business gets into it.
Being described well by ChatGPT is not a branding exercise. It is a sourcing problem, and the sources it prefers changed sharply in late 2025.
What changed
For most of ChatGPT's life, Wikipedia and Reddit dominated its citations, each accounting for something like 12 to 13 percent of them in the US.
Around September 2025 that broke. Tracking studies put Reddit's appearance rate falling from roughly 60 percent of responses to around 10, in a period that coincided with OpenAI's content licensing agreements and a visible shift toward professionally produced material. Forbes roughly doubled its citation share over the same stretch. Among news sources, PR Newswire recorded the largest gains of any domain.
Read that list again, because it is the whole strategy. What replaced forum threads was published journalism.
Why publishers win
The reasons are not mysterious. Publishers carry a long positive history in traditional search ranking, they apply editorial scrutiny before publishing, and they are timely in a way a static page is not.
An engine assembling an answer has to decide what to trust. A masthead with decades of search history and a named editor is a safer bet than a company's own About page, and the model behaves accordingly.
What this means for a business
Your own website is necessary and insufficient. It tells the engine what you say about yourself, which is the weakest category of evidence available.
What moves the needle is third-party published material: articles about you, on domains the engine already draws from, that state plainly who you are and what you do. That is not a trick. It is the same thing that persuades a human researcher, which is largely why it works.
The practical version
Be published where the engines already look. Make sure the facts in those articles are consistent, because contradictory sources produce vague answers. And accept that one article is a data point rather than a conclusion: engines weigh corroboration, so several consistent sources beat one good one.
The honest caveat
Nobody controls what a language model says. Citation behaviour shifts with licensing deals, model updates and the engine's own product decisions, as the Reddit collapse demonstrated in a single month.
What you can influence is whether good sources about you exist at all. If they do not, the engine will answer from whatever it can find, and that is the situation worth avoiding.
Getting started
Publisive places clients in the publications these engines already cite. We write the piece to your brief, you approve every word, and the article stays indexed and findable afterwards.
We write and place features in the publications our clients choose. About us
Frequently asked questions
01How do I get ChatGPT to cite my business?
By making sure accurate third-party published material about you exists on domains the engine already draws from, and that the facts in it agree with each other. Engines weigh corroboration, so several consistent sources do more than one good one.
02Why is my own website not enough?
Because it is self-reported, and engines treat it the way a journalist would: as what you say about yourself rather than as evidence. Third-party published coverage is a different category of source.
03Can anyone guarantee a ChatGPT citation?
No. Citation behaviour shifts with licensing deals and model updates, as the 2025 change demonstrated inside a single month. What you can influence is whether good sources about you exist at all.
