Advertising inside the next generation of search.

ChatGPT Ads are advertisements shown next to a conversation with the AI assistant. They are not part of the answer. Four signals together determine whether an ad is relevant: the context and intent of the conversation, the landing page, the title and ad copy and the context hints the advertiser provides. Context hints describe a situation and are not exact-match keywords. W69 AI Growth, based in Amstelveen (Amsterdam area), prepares campaigns for this and guides the onboarding as soon as access becomes available in the client's market.

Classic search ads begin with a term someone types. You pick keywords, you bid on those keywords and you write an ad to match. The entire campaign structure is built around isolated words and their variants.

A conversation with an AI assistant works differently. Someone first describes their situation, asks follow-up questions, has options lined up side by side, and only then arrives at a choice. By the time an ad can be relevant, far more is known than a single search term: the situation, the earlier questions and the direction the conversation is moving in.

That shifts the work. No longer expanding keyword lists, but describing the situations in which your offer is the right answer. And because the landing page counts as a relevance signal, that page is part of the campaign rather than the end of it.

A conversation almost always runs through the same stages, sometimes within minutes. If you have a message for each stage, you are not talking about purchase while someone is still orienting.

For each stage we record the underlying question, the message that fits it and the page that message points to. That is the AI Intent Map, and it is useful without media budget too: the same insights feed your organic AI visibility.

In a classic campaign you group keywords into ad groups. Here you group situations. A context cluster is a set of conversations that are about the same thing, with one promise, one piece of ad copy and one landing page.

The context hints you provide describe when that cluster is relevant: which topic it concerns, at which stage of the decision, and for what kind of customer. They steer how relevance is determined. They do not work like an exact-match keyword that guarantees an impression against a given term, and they do not buy you a position.

In practice that means the following. Instead of bidding on “crm software” you describe a cluster as: a small sales team that is not using the current system and needs an Outlook integration. The ad copy is about adoption, the landing page shows the integration, and the two reinforce each other as a signal.

As a result a campaign here is less a list and more a structure: a handful of well-described situations instead of hundreds of word variants.

Tools are arriving that build and manage your ChatGPT campaign automatically. That works, and for anyone who simply wants to switch on an advertising channel it is a perfectly good route. But an ad places you beside the answer. The answer itself remains untouched.

When an AI assistant names three providers and a single ad sits alongside, you want to appear in both. That takes two different kinds of work: a campaign structure for the paid side, and content, structured data, entities and authority for the organic side. W69 does both, from the same intent map, so that the message in the answer and the message in the ad do not contradict each other.

That is why a project with us does not start with the campaign, but with the question of which conversations your future customers are having before they become customers. See AI Advertising for the broad proposition and AI Visibility (GEO/AEO) for the organic side.

Go deeper