How do ChatGPT ads work? They buy the slot, not the answer?

ChatGPT ads are labelled placements shown beside ChatGPT's answer, and by OpenAI's own rules they cannot change a word of it. So the most valuable position on the screen, the brands the answer itself recommends, cannot be bought. Your ad can end up running next to a paragraph that names your competitor.

That is the part of ChatGPT advertising most brands have not priced in yet. The slot is for sale in more than 60 countries. The answer beside it is not for sale anywhere.

What exactly is for sale?

A sponsored unit next to the answer, shown to logged in adults on ChatGPT's Free and Go plans. OpenAI started testing in the United States on 9 February 2026, opened 31 European markets in August with self serve from 31 August, added Southeast Asia and Taiwan on 23 September, and now sells in more than 60 countries. Türkiye is on the self serve list.

The machine behind it is already large. OpenAI says ChatGPT ads passed a billion dollars in annualised revenue in under 200 days, with tens of thousands of advertisers. Buyers can bid on impressions, clicks or conversions, target by geography and custom audience, and track results with the OpenAI Pixel and a Conversions API.

Ads are matched to the topic of the conversation, the user's past chats and their past interactions with ads. When someone asks which CRM to buy, the ads that qualify are CRM ads.

What is not for sale?

The answer. OpenAI's ads principles say ads "do not influence the answers ChatGPT gives you" and are "clearly labeled as sponsored and visually separated from the organic answer". Every expansion announcement since February has repeated it.

Put that next to the matching rule and the consequence is plain. Your ad appears in exactly the conversations where ChatGPT is also writing its own recommendation for your category. If that recommendation names three competitors and not you, your ad is the sponsored footnote to someone else's endorsement. Budget cannot move you into the paragraph.

Who will never see your ad?

Two groups, and for many brands they are the ones that matter most.

Anyone on a paid plan. Plus, Pro, Business, Enterprise and Education carry no ads. In business software, professional services and most B2B categories, the person approving the purchase is often using ChatGPT through work.

Anyone in a sensitive category. During the test OpenAI kept ads away from health, mental health and politics.

For both groups the answer is the only placement that exists. A brand that sells to them and plans its ChatGPT presence around ads has no presence at all.

Why do ads make the answer worth more?

OpenAI's stated reason for advertising is to fund free and low cost access. More free usage means more conversations, and more buying decisions that start inside one. Ads do not compete with the answer for attention. They grow the audience that reads it.

And the scale of advertiser demand tells you something. Tens of thousands of businesses do not pile into a surface in six months unless buyers are already making decisions there. The organic answer is now the shelf every one of those advertisers would like to be on, and none of them can buy their way onto it.

What does the ad dashboard not show you?

Ads Manager reports impressions, clicks and conversions. OpenAI says advertisers receive only aggregate performance data. Nothing in it tells you what the answer beside your ad said, which competitors it recommended, or which sources it trusted to decide.

So a campaign can post a healthy click rate while the paragraph next to it sends the rest of the audience to a rival. That half of the screen is invisible from inside the ad account. It is the half Beaconn measures: for the questions your buyers actually ask, what ChatGPT answers, which brands it names, and which pages it cites to get there, alongside Gemini, Perplexity and Claude.

How should you plan ChatGPT ad spend?

Look at the answer first, then decide what the ad is for.

• The answer names you. The ad amplifies a position you already hold. Buy: this is the cheapest win on the screen.
• The answer names competitors, not you. The ad runs as a footnote to a rival's endorsement. Fix the sources the answer relies on first, then buy, or run both together and watch the answer move.
• The answer names no one clearly. The ad fills an open shelf. Buy, and publish to take the answer while it is uncontested.
• Your buyers are on paid plans or in a sensitive category. The ad cannot run. The answer is your only channel, so put the budget into it.

Every row starts with the same step: knowing what the answer says today, question by question. Without it, ad spend is being allocated blind.

Frequently asked questions

When did ChatGPT start showing ads? OpenAI began testing ads in the United States on 9 February 2026. By 23 September 2026 they were available in more than 60 countries.

Which ChatGPT plans show ads? Free and Go. Plus, Pro, Business, Enterprise and Education plans are ad free.

Can ChatGPT ads change what ChatGPT recommends? No. OpenAI says ads do not influence the answers, and are labelled and shown separately from them. The brands the answer recommends cannot be bought.

Can a business in Türkiye buy ChatGPT ads? Yes. Türkiye is listed as available for self serve in OpenAI's Ads Manager, provided the advertising entity is based there.

How do I know what ChatGPT says about my brand before I buy ads? Ask the questions your buyers ask and read the answers, engine by engine. Beaconn does this across ChatGPT, Gemini, Perplexity and Claude and shows which brands each one names and which sources it relies on.