What a ChatGPT ad is
When someone asks ChatGPT for a recommendation, a comparison or a “best X for Y”, OpenAI can show a clearly labelled sponsored result alongside the answer. The placement is matched on the meaning of the conversation, not on a keyword, which means your product information has to be structured for a model to understand rather than for a keyword to match.
Why the intent is different
A Meta impression interrupts someone scrolling. A Google click comes from a typed query. A ChatGPT placement appears inside a conversation where the person has already described their problem in detail: skin type, budget, roof size, the occasion. That context is what makes the channel convert when the match is right and waste money when it is not.
What it costs
Early inventory is priced below comparable Google Search intent because few advertisers are bidding. We quote a test budget per account rather than publish a rate, because the numbers move monthly as more categories open. Expect the same pattern as Google Shopping a decade ago: cheap, then crowded.
Who should run it now
- Brands already profitable on Meta or Google, because the conversion data from those accounts is what tells you whether OpenAI is working.
- Products with clear, factual descriptions and a known price. Ambiguity hurts a model-matched placement.
- Categories with research-heavy buyers: skincare, supplements, equipment, travel, solar, education.
How to prepare
- Product data. Plain-language descriptions of what the product is, who it is for, what it costs and what it is not. Feeds built for keyword matching are not enough.
- Site content. Pages that answer the questions people ask a model: “is this suitable for sensitive skin”, “how long does delivery take”, “what is the minimum order”. These help the ads and get you cited organically.
- Tracking. OpenAI placements need to land in the same reporting as Meta and Google so they are judged on profit, not novelty.
We run OpenAI ads as a third engine for qualifying clients. See the ChatGPT ads service page and the sports nutrition case study, where it was added in month five.