ChatGPT Ads and GEO: What Applies in Germany Now

OpenAI initially launched the advertising test in the United States on February 9, 2026. After adding further markets, August brought the largest expansion to Europe so far.

OpenAI has announced plans to expand self-service access in Europe. On the latest official availability page, however, Germany is not currently listed as a self-service market. Anyone planning to create a campaign directly in Ads Manager should therefore check the status again before launch. Independently of self-service access, OpenAI lists the Ads Solutions Team as well as agency and technology partners as ways to access advertising in European markets.

This is an important distinction from the idea that a brand can simply buy its way directly into a ChatGPT answer. According to OpenAI, that is not possible.

An ad unit can include, among other things:

• the advertiser’s company name
• a favicon
• a headline
• a description
• a landing page
• an image

To decide which ad to show, the system considers, among other things, what the conversation is about and what intent can be inferred from it. Information from the ad and landing page is also taken into account.

Advertisers can also provide so-called Context Hints. These describe topics, conversations, or terms for which their offer may be relevant.

At first, that sounds similar to keywords in Google Ads, but it does not work the same way. OpenAI explicitly states that Context Hints are not exact-match keywords and do not guarantee delivery within a particular conversation.

That makes ChatGPT Ads less predictable to plan than traditional search ads. At the same time, this is exactly what makes the channel distinctive: an ad is not shown only for one search term, but can appear during a longer research or decision-making process.

That distinction matters. The $3 to $5 range does not mean businesses in Germany will pay that amount on average for every click.

OpenAI uses a relevance-weighted second-price auction. In addition to the bid, the system therefore considers how relevant an ad is expected to be in the specific situation.

Three basic campaign objectives are currently available:

• CPM for reach and impressions
• CPC for clicks and traffic
• oCPC for optimizing toward a downstream conversion

With oCPC, the campaign is optimized toward a defined conversion event. Billing still occurs for valid clicks rather than per conversion.

For conversion-optimized campaigns, OpenAI currently does not specify a recommended bid.

Measurement has also become much more developed than it was at the start of the advertising test. Ads Manager now provides metrics including impressions, clicks, spend, CTR, average CPC, CPM, and conversions.

Conversion tracking can be implemented through the Conversions API or a JavaScript pixel. Static parameters such as UTM tags can also be passed to landing pages.

This makes it possible to evaluate ChatGPT traffic in existing analytics systems. For businesses, that matters more than any early market benchmark. A new advertising channel can only be tested properly if you can see what happens after the click.

For businesses, the practical difference is that a Google Ads campaign can usually be planned in considerable detail using existing search volumes, historical CPCs, and keyword data.

ChatGPT Ads still lack much of that historical experience.

That does not automatically make the channel worse. It simply means early campaigns should be treated more like controlled experiments. Anyone trying to make a universal judgment about the economics of ChatGPT Ads after only a few days is working with a data foundation the market does not yet meaningfully provide.

For me, this is the most important point in the entire discussion.

Paid visibility and organic AI visibility answer two completely different questions.

With ChatGPT Ads, the question is: How can I place a clearly labeled ad in front of a relevant user?

With GEO, the question is: What does an AI system need to find and understand about my company for my brand to be considered a relevant answer to the right question?

Despite that technical separation, the two disciplines can still meet within the customer journey.

A user might see an ad first and later ask ChatGPT which providers could solve their problem. Or they may research alternatives and ask how different companies compare.

At that point, what ChatGPT knows organically about those companies suddenly matters.

But that is not a ranking effect caused by the ad. It is simply a second interaction with the same brand.

For the paid side, I would clarify four things first:

• Does your target audience actually use ChatGPT for research, comparisons, or purchase decisions?
• Does the landing page fit the question and context that might lead a user to click the ad?
• Are conversion tracking and analytics working before campaign spend begins?
• Is an inquiry or sale valuable enough to justify testing a channel that is still relatively new?

Only after that would I release budget.

Organic visibility requires a different kind of work.

The goal is to answer the questions your audience actually asks, support claims with credible evidence, make real experience visible, and position your brand clearly for both people and machines.

That work falls under Generative Engine Optimization.

Technical accessibility is part of it as well, but it is only the foundation. A technically perfect website is not automatically recommended in ChatGPT if the decisive information is missing or interchangeable.

The biggest mistake would therefore be to wait for ChatGPT Ads and ignore organic visibility in the meantime. Advertising can be activated relatively quickly. Credible content, strong sources, and a clear topical profile take much longer to build.

Before running a test, I would answer these four questions:

• Does my target audience already use ChatGPT to research complex decisions?
• Is my offer complex enough that conversational context could provide an advantage?
• Is the economic value of a strong click or lead high enough?
• Can I measure conversions reliably?

If several answers are “no,” I would not reallocate budget simply because of the current hype.

If the conditions are right, however, a limited test can be worthwhile. Precisely because the channel is still young, the goal should not be to prove a new performance channel immediately. It is more useful to learn which conversations, messages, and landing pages actually work.

That is why I would measure and optimize the two areas separately.

Paid answers the question of how a company can become visible through budget.

GEO answers the question of whether a company appears even when nobody is paying for the mention.