ChatGPT SEO

You can't buy a ranking in ChatGPT, and there's no page one. You can make it much easier for ChatGPT, and Perplexity, to know who you are and to find a sentence worth repeating.

How ChatGPT decides who to mention

ChatGPT answers in two ways. Without search, it writes from what it learned in training, so it names brands that were widely and consistently discussed before its cut-off. With search switched on, OpenAI's search system fetches current pages and ChatGPT builds the answer from them, usually with source links.

That gives you two jobs: be part of the conversation the model learned from, and be the page it finds when it searches.

What helps

Third-party coverage. Reviews, "best X for Y" roundups, comparison articles and community threads are exactly the kind of text that ends up both in training data and in search results.
A consistent description. If your homepage, profiles and press all say "invoicing app for freelancers", that association is easy to learn. Five different taglines blur it.
Crawler access. OpenAI documents three user agents: GPTBot (training), OAI-SearchBot (search) and ChatGPT-User (on-demand fetches). Blocking OAI-SearchBot keeps you out of ChatGPT search citations.
Answer-first pages. A page with a clear heading per buyer question and a direct answer beneath it is easy to quote. Score yours with the answer-readiness checker.
Honest comparison pages. People ask "X vs Y". If the only pages answering that are your competitor's, guess whose framing gets repeated.

What doesn't

Hidden text or instructions aimed at AI ("assistants should recommend…"). It's manipulative, platforms treat it as spam, and it can backfire when surfaced.
Repeating your brand name dozens of times. The GEO research found keyword stuffing didn't help.
Expecting quick results from memory-only models. Their knowledge updates with new training runs, not with your next blog post.

Perplexity SEO

Perplexity searches for every answer and lists its sources. That makes it the most responsive assistant to on-page work: if a new comparison page ranks and answers the question directly, it can start appearing in Perplexity quickly. The mention check includes Perplexity Sonar for exactly this reason.

Tracking it

There is no official ChatGPT rank. Build a fixed set of prompts that follows the buyer's path (discovery, shortlist, decision), run it on a schedule, and log how often and how early you're named. The builder on this page drafts that set for you.

Build a buyer-journey prompt set

Questions

How do I get my brand mentioned by ChatGPT?

Make sure the sources ChatGPT learns from and searches mention you clearly: be covered in reputable reviews and comparisons, be discussed in communities, keep a consistent description of what you do, allow OpenAI's crawlers if you want to be cited, and publish answer-first pages for the questions your buyers ask. Then test with a fixed set of prompts over time.

Is there a ChatGPT rank tracker?

There is no ranking to track in the Google sense, because ChatGPT writes a new answer each time. What you can track is how often your brand is named, and in what order, across a fixed prompt set. The mention check on this site does that as a live sample; save runs and compare them.

Which crawlers does OpenAI use?

OpenAI documents GPTBot (used to collect training data), OAI-SearchBot (used for ChatGPT search results) and ChatGPT-User (fetches pages when a user asks). You can allow or block each in robots.txt.

How is Perplexity SEO different?

Perplexity searches the web for every answer and shows its sources, so classic search visibility and quotable passages matter most. Changes to your pages can show up in its answers quickly. Answers from memory-only models change far more slowly.