These days, when people make a purchasing decision, they’re increasingly turning to ChatGPT. Not Google. Not a store. But a chatbot. According to OpenAI, about 20 percent of all ChatGPT queries have a commercial intent. That’s sparking interest. In February 2026, OpenAI launched its first advertising formats in ChatGPT. At the Cannes Lions, the company aggressively courted brands and agencies.
Für Marketingverantwortliche stellt sich damit eine neue Frage. Wie wird in der KI-Suche künftig geworben? Und was bedeutet das für die eigene Sichtbarkeit? Dieser Beitrag ordnet die Entwicklung ein.
Key Points at a Glance
- ChatGPT Ads are labeled ad spaces that appear below the AI response. They target the context of the conversation, not individual keywords.
- OpenAI has been testing these formats in the U.S. since February 2026. According to OpenAI, about 20 percent of the inquiries indicate a clear intent to purchase.
- In-chat purchases came to a standstill in 2026. OpenAI scaled back its Instant Checkout feature.
- Google is catching up in the areas of AI advertising and AI commerce. Competition among platforms is just getting started.
- Seit August 2026 sind ChatGPT Ads auch in Deutschland verfügbar. Für Unternehmen zählt jetzt: die KI-Suchintention verstehen, bevor Budget fließt.
What are ChatGPT Ads?
ChatGPT Ads are paid placements within ChatGPT. They appear as a box below the actual answer. They are clearly labeled “Sponsored.” OpenAI thus clearly separates advertising from editorial content. These formats have been in testing since February 2026. For now, they are visible only to registered adult users of the free tier and the affordable Go tier in the U.S.
The key difference lies in targeting. Traditional Google ads respond to a keyword. ChatGPT Ads respond to the context of the conversation. The AI recognizes the topic, the intent, and the depth of the search. Only then does a dedicated system check whether a suitable ad exists. Here’s an example: Someone is planning a hiking trip in ChatGPT. An outdoor gear retailer might appear as a relevant result—not because the word “hiking boots” was mentioned, but because the conversation is heading in that direction.
Wichtig ist die Reihenfolge. ChatGPT formuliert zuerst seine Antwort. Erst danach entscheidet ein getrenntes System über eine Anzeige. Die Werbung beeinflusst die Antwort also nicht. Das ist ein zentrales Vertrauensargument. Es unterscheidet die KI-Antwort vom klassischen Suchergebnis, bei dem Anzeigen oft ganz oben stehen. Wie sich KI-Antworten generell von der klassischen Suche unterscheiden, erfahren Sie in unserem Beitrag „lohnt sich SEO noch?“ Anfangs lief der Test nur in den USA. Seit August 2026 ist ChatGPT Ads auch in Europa und Deutschland verfügbar. Anzeigen sehen weiterhin nur Nutzer der Free- und der günstigen Go-Stufe.
Insights from the U.S.: The First AI Ad Formats
In June 2026, OpenAI made its debut at the Cannes Lions. There, the company pitched its new advertising product to brands and agencies. The message: The advertising industry is shifting from an “attention economy” to an “intelligence economy.” Users arrive with a specific task. David Dugan, who heads OpenAI’s global advertising business, called them “super intentional.” That, he said, is precisely what makes the context valuable for brands. This is according to a report by Marketing Dive.
OpenAI places the ads in the middle of the funnel—that is, where users compare products and prepare to make a decision. According to Dugan, about 20 percent of all ChatGPT queries have a direct commercial intent. According to Marketing Dive, ChatGPT has nearly one billion weekly users. That’s an enormous reach for a platform that’s only four months old.
Initial metrics are also available. The industry publication eMarketer reports on an important trend: Since its launch in February, the number of ads clicked away from has dropped by 50 percent. This suggests that the ads are more relevant. OpenAI is rapidly expanding its infrastructure. A self-service ad manager has been available in the U.S. since April. Its structure is reminiscent of Google Ads. Partners such as Criteo and StackAdapt provide familiar workflows. According to Marketing Dive, more than 2,000 brands are already advertising on ChatGPT via Criteo.
When it comes to spending, the format remains exclusive. According to eMarketer, the minimum bid is $50,000. Through Criteo, tests are possible starting at $10,000. Providers report click prices ranging from $2.50 to $8.00. By comparison, in traditional Google Search, CPCs often range from $1 to $3. These figures come from the service provider Ryze AI, which markets AI-powered ads. They should therefore be viewed from the provider’s perspective, not as a neutral measurement.
Three formats are emerging. First, sponsored product recommendations displayed as a card below the answer. Second, interactive extensions that allow users to ask the brand follow-up questions. Third, editorial resource cards for complex B2B topics. What has worked so far is a distinctive tone. An ad is more effective when it sounds like a recommendation—not like a banner. OpenAI has set a clear goal for its advertising business. According to eMarketer, it is expected to generate approximately $100 billion in revenue by 2030. This demonstrates just how ambitious the company is.
For Context
OpenAI has outlined four guiding principles for its advertising business. Ads remain separate from user responses. Conversation content is not shared with advertisers. Users retain control and can delete their history. And the platform is designed to deliver long-term value. That sounds reasonable. Whether it will hold up over the long term, however, will only become clear under economic pressure. An advertising company is judged by its revenue, not by statements of intent.
From Search Results to Purchase
Advertising is only part of the story. The bigger game-changer lies in direct purchases. To that end, OpenAI launched Instant Checkout in September 2025. In February 2026, the company made the feature available to all U.S. users. The idea: find products and buy them immediately in the chat—without having to go through a store. Technically, this is powered by the Agentic Commerce Protocol, which OpenAI developed in collaboration with the payment service provider Stripe.
The reality turned out to be more sobering than the vision. As early as March 2026, OpenAI rolled back Instant Checkout. According to CNBC, very few merchants went live with the feature. Industry reports cite fewer than 30 out of millions of Shopify merchants. The reasons are instructive. Product data from the web was often out of date. Inventory, shipping, and taxes were difficult to track. And while users did research in the chat, they ended up buying elsewhere.
OpenAI is now shifting the purchasing process to its own shopping apps. ChatGPT is focusing more on product search and comparison. This is a notable step back. It shows that seamless AI-powered purchasing is technically challenging. The headlines got ahead of reality. This is exactly where a level-headed assessment is needed. Hype and market readiness often don’t align.
Google is taking advantage of this gap. According to an analysis by Fast Company, real-time prices and inventory data are already live on Gemini. Purchases are made directly in the chat via Google Pay. Major brands such as Gap and Ulta Beauty are among the first to enable the entire purchase process to take place within the conversation. At the same time, a standards battle is unfolding. OpenAI and Stripe are backing their Agentic Commerce Protocol, while Google, Shopify, and an alliance of retailers are supporting the competing Universal Commerce Protocol. Amazon, on the other hand, is stonewalling. The company is blocking third-party AI agents on its site.
How might the AI advertising market evolve?
From today's perspective, several lines of development are emerging. They are not mutually exclusive. They are likely to run parallel to one another.
1. Direct purchases in chat
That’s the most radical vision. AI finds the product and completes the purchase. For retailers, this shifts the nature of competition. It’s not the most attractive store that wins, but rather the best product data and the best availability. Anyone who wants to be visible here needs clean, structured data. However, the setbacks of 2026 show that it will still be some time before this technology is ready for mass adoption.
2. Native Advertising in the Response
This scenario is already playing out. Sponsored posts appear below the AI’s response. They come across as recommendations. That’s exactly what makes them effective—and at the same time, tricky. Because the line between advice and advertising is blurring. A new discipline is emerging for brands. Ads must blend into the flow of the conversation. Anyone who advertises too blatantly will be clicked away.
3. Traditional Advertising Spaces Related to AI
A pragmatic approach is also still possible. AI providers could place ads in the margins and within apps—similar to how it works today with search engines and social media. Here, familiar models such as cost per click and reach are key. For many advertisers, this would be the easiest way to get started.
4. Competition Among Multiple Platforms
The market structure is key. OpenAI isn’t alone. Google is pushing ahead with Gemini ads. Microsoft is integrating advertising into Copilot. Anthropic is exploring commercial features. For advertisers, more competition is a good thing. It drives down prices and increases choice. The big question is: Will an open market emerge? Or is a monopoly simply shifting from search to chat? This is a question worth keeping a close eye on.
ChatGPT Ads ist jetzt in Europa verfügbar
Am 18. August kündigte OpenAI den Start von ChatGPT Ads in 31 europäischen Märkten an. Deutschland und Österreich gehören dazu. Seit dem 31. August ist der Self-Service über den Ads Manager offen. Interessierte Unternehmen registrieren sich unter ads.openai.com. Wie in den USA sehen nur Nutzer der Free- und Go-Stufe die Anzeigen. Plus, Pro und Enterprise bleiben werbefrei.
Ein kurzer Blick in den Ads Manager
Der Einstieg ist bewusst einfach gehalten. Der Ads Manager schlägt zunächst einen empfohlenen Plan vor. Grundlage sind Ihre Website und Ihr Unternehmen. Sie prüfen die Vorschläge und passen sie an. Festlegen lassen sich der Name des Werbetreibenden und das Ziel, etwa Klicks. Dazu kommen die Standorte, die Ziel-URL und die Gebotsstrategie. Ein Beispiel ist die Strategie „Maximale Ergebnisse“. Auch ein durchschnittliches Tagesbudget geben Sie an, im Test etwa 100 US-Dollar. Eine Vorschau zeigt die spätere Anzeige. Die Logik erinnert stark an Google Ads. Wer damit gearbeitet hat, findet sich schnell zurecht.
Tipps für den Start
Ein guter Start ist überlegt, nicht überstürzt. Diese Punkte helfen Ihnen dabei.
- Klein starten. Wählen Sie ein klares Ziel und ein moderates Tagesbudget. Testen Sie erst, dann skalieren Sie.
- Eigene Zielgruppen nutzen. Sie können benutzerdefinierte Zielgruppen anlegen, auch auf Basis eigener Kundendaten wie E-Mail-Listen. Achten Sie dabei streng auf DSGVO und Einwilligung.
- Suchintention zuerst verstehen. Prüfen Sie vor dem ersten Euro, was Ihre Zielgruppe ChatGPT wirklich fragt.
Genau hier setzt unsere Prompt-Recherche an. Wir analysieren, mit welchen Prompts Ihre Zielgruppe in der KI-Suche unterwegs ist. So erkennen Sie, welche Fragen, Vergleiche und Formulierungen wirklich zählen. Auf dieser Grundlage planen Sie Anzeigen und Inhalte, die zur echten Nachfrage passen. Das schützt Ihr Budget vor teuren Fehlversuchen, statt blauäugig zu starten. Sprechen Sie uns dazu an.
What does this mean for German companies?
Für kleine und mittlere Unternehmen heißt das aber nicht, sofort Budget zu verschieben. Klüger ist ein überlegter Testeinstieg. Machen Sie sich mit dem Ads Manager vertraut. Prüfen Sie in Ruhe, welche Anzeigenformen zu Ihrer Strategie passen.Online-Shops sollten die technischen Voraussetzungen schaffen, um direkte Käufe im Chat zu ermöglichen, auch wenn es diese Funktion noch nicht gibt. B2B-Unternehmen müssen sich überlegen, wie sie mit relevanten Inhalten auf KI-Suchen reagieren können, denn klassische Werbeflächen mit generischen Botschaften wie „Kaufen Sie unsere XY-Maschine“ werden in der KI-Suche kaum überzeugen.
Even more crucial for SMEs than paid ads is organic visibility in AI systems such as ChatGPT, Gemini, or AI Overviews. Any business that appears as a trusted source here has a clear competitive advantage. This requires: a clear strategy, technically sound websites, and AI-optimized content with structured data and a logical content architecture.
A well-designed customer journey that takes all online channels into account remains essential. AI often engages users as early as the initial research phase, but traditional touchpoints such as ads, social media, and organic search continue to be crucial.
For now, data remains scarce, as AI providers offer fewer insights than Google or Meta. Therefore, anyone who wants to understand traffic from AI must rely on accurate monitoring and their own analyses.
Tip: Don't blindly trust the numbers provided by the platforms. Establish your own metrics and stay agile!
An Inconvenient Truth
ChatGPT Ads are powered by the same mechanisms as those used by the major U.S. platforms. According to eMarketer, OpenAI expanded the sharing of limited user data with advertisers in May 2026. So anyone working with AI advertising is once again handing over data to a U.S. corporation. Carefully review data protection and GDPR requirements before reallocating budgets. And maintain a healthy dose of skepticism. A new platform isn’t automatically a better platform—at first, it’s just a new dependency.
Tip: Understand AI search early on, invest later
The AI advertising market is just emerging. The first ad formats are live, reach is extensive, and interest is growing. At the same time, 2026 reveals its limitations. Direct purchases via chat are stalling. Measurability is lacking. And a new monopoly cannot be ruled out. Those who blindly pour budgets into AI ads now will learn the hard way. Those who ignore the issue will miss out on a major development. The right approach lies somewhere in between.
Unsere Empfehlung ist klar. Bauen Sie zuerst Ihre organische KI-Sichtbarkeit auf. Schaffen Sie Inhalte, die KI-Systeme gerne zitieren. Halten Sie Ihre Produktdaten sauber. Wenn Sie Budget übrig haben, testen sie zuerst mit geringen Budget diesen Kanal. So sichern Sie sich den Vorsprung, ohne das Risiko der ersten Stunde zu tragen.
Would you like to prepare your company for ChatGPT, Gemini, and Google AI Overviews today?
Paseo Marketing, based in Karlsruhe, provides honest, peer-to-peer consulting to German companies, small and medium-sized businesses, and B2B "hidden champions." We don't shy away from addressing uncomfortable issues and work with you to implement solutions. Request your free initial consultation now!
Alexander Keller
Chief Search Architect
kontakt@paseo-marketing.de
0721 – 911010
