ChatGPT SEO: how to be the source it names

There is no ranking to climb inside ChatGPT. When it answers a question with current information it retrieves a handful of pages, synthesises them, and names some of them. Your entire goal is to be in that handful.

What has no effect

• Keyword density. There is no keyword matching step to game.
• Publishing frequency. A model has no notion of how often you post.
• Paying anyone. Citation inside a synthesised answer is not a placement you can buy.
• Meta keywords, hidden text, and every other trick that stopped working on search engines years ago.

What does

Being retrievable comes first: if a page cannot be crawled and understood by conventional search infrastructure, it will not reach the retrieval step at all. Answer engines lean heavily on the same open web everything else does.

After that, it is a question of whether your page is the safest thing to quote. Safety here means: the claim is attributed, the organisation making it is identifiable, the statement is specific enough to be useful, and something outside your own domain corroborates it.

Measuring it

No engine publishes a report on how often it quoted you, so measurement means asking. Take the questions your buyers actually ask, put them to the engine on a schedule, and record whether it named you, a competitor, or nobody.

Over weeks that produces a citation share — the percentage of your important questions where you are named. It is the only number in this field that behaves like a rank position: comparable over time and attributable to work. [internal: the full GEO guide]