- Adthena says it is tracking more than 50,000 daily ChatGPT ad placements across 600+ advertisers, revealing early patterns in how brands are writing ads for conversational search.
- The data suggests ChatGPT ads need shorter, calmer and more context-aware copy than traditional Google Ads, because users are already inside a decision-making conversation.
ChatGPT ads are beginning to develop their own creative rules, and they do not look like a simple copy-and-paste version of Google Ads.
In a new analysis of more than 50,000 daily ChatGPT ad placements, Adthena said it is tracking ads across 600+ advertisers and seeing clear patterns in the way early advertisers are writing for the format.
The biggest takeaway: ChatGPT ads are not written for someone typing a keyword into a search box. They are written for someone already inside a conversation, comparing options and moving toward a decision.
ChatGPT ads are built around conversational intent
OpenAI describes ads in ChatGPT as sponsored messages that appear below relevant conversations. The ad unit includes an advertiser name, favicon, headline, description, landing page and image asset.
OpenAI says ads are shown to Free and Go users in supported markets, while Plus, Pro and Business users do not see ads. The company also says ads are separate from ChatGPT’s answers and are clearly labeled as sponsored.
That separation matters. Advertisers are not buying their way into the answer itself. They are buying placement near a moment of intent, after the user has asked ChatGPT for help, comparison or direction.
Brand names are leading the ad copy
One of the clearest patterns in Adthena’s data is that many top ads lead with the brand name.
Adthena describes the dominant format as a “Brand: Benefit” structure, where the brand name appears first, followed by a short value claim. That is different from the way many Google Ads headlines are written, where advertisers often start with the keyword, the problem or the offer.
In ChatGPT, the brand may be doing more of the work. If a user is already comparing options, the ad does not need to recreate the full search query. It needs to make the brand feel like a logical next step in the conversation.
The copy is much shorter
ChatGPT ad copy is also much tighter than traditional search copy.
According to Adthena’s analysis, ChatGPT ad headlines average about 30 characters, with an observed ceiling around 36 characters. Body copy averages around 19 words, usually structured as two short sentences.
That creates a different writing constraint. Google Ads gives advertisers more room across multiple headlines, descriptions and assets. ChatGPT ads appear in a more compact conversational environment, where long or loud copy can feel out of place.
For advertisers, that means every word has to carry weight. A vague benefit line or generic phrase like “discover the difference” is less useful when the format only allows a small amount of text.
Context mirroring may be the real unlock
The most interesting part of the format is context.
Adthena says some of the strongest ChatGPT ads mirror the user’s prompt directly. For example, an ad for running shoes may reference a user’s stated transition from a 5K to a half marathon, rather than using a generic running-shoe message.
That points to the strategic difference between search ads and conversational ads.
Google Ads are often built around keyword intent. ChatGPT ads can be built around the broader context of what the user is trying to accomplish. The ad does not only match a query. It can feel like it belongs to the conversation.
Calm, specific copy appears to fit the format better
Adthena’s findings also suggest that ChatGPT ads are quieter than many traditional performance ads.
Numbers, dollar amounts, trial lengths and specific proof points show up frequently. Exclamation marks and hype-heavy phrasing are rare, according to the analysis.
That makes sense for the environment. ChatGPT’s interface is calm and answer-driven. An ad that suddenly sounds like a loud display banner can break the tone of the experience.
In conversational search, trust may depend on sounding helpful rather than promotional.
Why Google Ads copy may not transfer well
The instinct for many advertisers will be to reuse Google Ads headlines and descriptions.
That may be a mistake.
Google Ads are written for users actively searching. ChatGPT ads are placed near users who may already be evaluating, narrowing or deciding. The user has often explained more context than a keyword can capture.
That changes the job of the ad. It has to be shorter, more specific and more aligned with the conversation. It should feel like a relevant next step, not an interruption.
Adthena has also launched ChatGPT Ads Intelligence, a product designed to track ad presence, competitor visibility, prompts that trigger ads and ad copy across ChatGPT placements. The company says it monitors more than 300,000 daily prompts in real time.
Measurement is still early
OpenAI’s own ad reporting is still limited compared with mature paid search platforms.
Adthena argues that advertisers need competitive context, prompt-level visibility and share-of-search style tracking because native reporting does not yet show the full landscape. In its ChatGPT Ads Intelligence product page, Adthena says it tracks which brands are advertising, which user questions trigger ads and how share of search compares with competitors.
That gap matters because a new ad channel can look simple at first. Impressions and clicks are useful, but they do not tell advertisers which prompts they are missing, which competitors are showing up or how the market is forming around different intent clusters.
The Query Post view
ChatGPT ads should not be treated as another Google Ads placement.
The early data suggests a different creative logic: lead with the brand, keep the message short, use concrete proof and match the user’s conversational context. That is a very different skill from writing three Google Ads headlines around a keyword.
The bigger shift is that paid search is moving from query matching toward intent matching. A keyword tells advertisers what the user typed. A conversation tells advertisers what the user is trying to do.
That makes ChatGPT ads potentially powerful, but also harder to manage with old habits. Repurposed Google Ads copy may fit the character limits, but still miss the moment.
For advertisers, the practical move is to start treating ChatGPT as a separate channel now. Track where competitors appear, study which prompts trigger ads and write creative that sounds native to a conversation rather than a search results page.
The first prompt may not replace the first click. But it is becoming a new place where brand preference starts to form.
