What we do on ChatGPT

How does a brand appear in ChatGPT?

A brand appears in ChatGPT in two separate ways: through paid placements clearly labelled as ads, or organically, by being mentioned as a source or a recommendation in the model's answer. The relationship between the two is similar to the one between search ads and SEO in classic search. One buys visibility; the other requires being a trustworthy source.

On this page we treat ChatGPT both as an advertising platform and as a visibility surface. The same approach largely applies to other AI assistants such as Gemini, Perplexity, Copilot and Claude. The differences show up in how each platform chooses its sources and in its advertising policies.

Are there ads in ChatGPT, and are they available in Türkiye?

OpenAI is testing ads in ChatGPT gradually and in a limited set of markets; access for advertisers in Türkiye varies and may change over time. That is why we do not give a specific launch date, pricing model or list of countries. We check the situation against the platform's latest announcements at the start of every engagement.

According to the framework OpenAI has described, ads are shown in a space that is separate from the answer itself and labelled as advertising, and the company stresses that ads do not influence the model's answers. Rules such as which user groups and subscription types will see ads, and which topics will be kept closed to advertising, are also shaped by the platform's decisions and can be updated.

In practice this means that, for many brands today, ChatGPT advertising is not a budget line but something to prepare for. It makes sense to build the groundwork now so that the first test can be run quickly and measurably once access opens. We cover the wider picture of paid placements in generative engines separately on our GEM page.

Which type of visibility do we choose, and when?

Visibility in ChatGPT and similar assistants is not a single product; different surfaces do different jobs. The table below summarises the types we work on today or are preparing for. Because platforms are developing these surfaces gradually, formats may change over time.

All the organic surfaces in the table draw on the same foundation: a crawlable site, content that gives clear answers and consistent information about your brand. A paid placement does not replace this foundation. An ad built on a weak foundation leaves the user who clicks without content that can persuade them.

What is the difference between a ChatGPT ad and a Google search ad?

The fundamental difference between a ChatGPT ad and a Google search ad is the context the ad accompanies: a search ad sits alongside a short query, whereas an ad inside a chat often accompanies a long, detailed conversation. The user has already described their needs, expectations and constraints in their own words. This calls for ad copy that is more explanatory and better suited to the context.

Keyword list logic alone is not enough here; you need to think in terms of topic, intent and the stage the conversation has reached. The search term reports and conversion data from your existing search campaigns are the most valuable resource for this preparation, because they show which questions are closest to a sale.

What do we do to become a source in ChatGPT answers?

Being cited as a source in ChatGPT answers requires the model to recognise your brand, be able to reach your pages and find your content quotable. We call this work GEO, or generative engine optimisation. GEO is the combination of content, technical and reputation work that helps a brand appear with accurate information, and as a source, in the answers of AI assistants.

• Accessibility: Managing AI crawlers' access to the site deliberately through robots.txt and server settings, and making sure important content is not buried in areas that only load through scripts.
• Quotable structure: A direct answer to the question at the top of every important page, clear definition sentences, comparison tables and a last-updated date.
• Entity consistency: The brand name, field of activity, address and service description appearing the same way on the site, in directories and on social profiles, supported by structured data.
• Third-party signals: The brand being mentioned naturally in industry publications, comparison content and user reviews.
• Question map: Identifying the questions your customers ask assistants and making sure there is a page that answers each one.

Most of this work shares the same ground as classic SEO. The difference is that the goal is not a ranking, but being mentioned in the right context inside the answer. You can find a detailed definition of the concept in our article on what GEO is.

How can you prepare for ChatGPT ads now?

Preparation means building in advance the core components that will let you launch a measurable test on the day access opens. All of these components are useful in other channels today as well, so the preparation is not wasted effort even if access is delayed.

1. 1.Question and intent map: Grouping the questions users ask assistants about your product or service by intent. This map sets both the ad contexts and the priorities for organic content.
2. 2.Measurement foundations: Defining form, call, basket and sale events correctly, with server-side measurement and a CRM connection where possible. A new channel's contribution can only be distinguished if this groundwork is in place.
3. 3.Creative that suits conversational language: Short, explanatory copy that completes the user's question instead of a slogan. The ad needs to read like a natural continuation of the conversation.
4. 4.Landing pages matched to questions: Users arriving from an assistant are usually in the middle of their research; send them not to the home page but to the page that picks up where their question left off.
5. 5.Test budget and decision criteria: A test share set aside without putting the main campaigns at risk, and the criteria for judging the test a success written down in advance.

What do we pay attention to in sensitive sectors?

In health, finance, law and products aimed at children, AI visibility requires extra care. Platforms may restrict advertising on sensitive topics. On the organic side, a model associating incorrect or outdated information with your brand creates a serious reputational risk.

In these sectors we review content with you for regulatory compliance, avoid wording that makes medical or financial claims, and make expert and author information visible. We regularly monitor the answers assistants give about your brand and, when we notice incorrect information, we work to find its source and get it corrected.

How is ChatGPT visibility measured?

ChatGPT visibility is measured by putting a defined set of questions to assistants at regular intervals and recording whether the brand is mentioned in the answers, in what context and with which source. Because the same question can get different answers at different times, what matters is the trend over time, not a single screenshot.

• Whether the brand is mentioned, and the context in which it appears alongside competitors
• Which of your pages are shown as sources in answers
• Referral traffic from AI assistants in analytics, and the conversions from that traffic
• The accuracy and currency of the information assistants give about your brand

Once the advertising side opens, we bring impression, click and conversion data together with these organic indicators in the same report. That way we can tell whether a paid placement is replacing organic visibility or complementing it. We also track Google's AI surfaces within the same measurement framework through our AIO work.

What are the common mistakes with ChatGPT visibility?

• Blocking AI crawlers without realising it, or never considering which content is open to them.
• Focusing only on your own site and neglecting how the brand is described in other sources.
• Using hidden text, fake reviews or mass-produced content in an attempt to steer assistants; these methods carry both reputational and policy risks.
• Allocating budget once ad access opens without a measurement set-up, and judging the outcome on clicks alone.