OpenAI is expanding advertising across ChatGPT, but marketers using the platform say its campaign tools remain limited compared with those available from Google and Meta.

The question is not only whether OpenAI can reproduce the controls offered by established platforms. ChatGPT matches advertisements to the context of a conversation, creating different requirements for targeting, measurement, and advertiser reporting.

The company began testing advertisements in February and has since added self-service campaign management, cost-per-click bidding, conversion tracking, and location targeting. An OpenAI spokesperson told Business Insider that ChatGPT ads were appearing in seven countries at the time of its report.

OpenAI’s advertising business had reached $100 million in annual recurring revenue by March, according to Business Insider. The company has also projected about $2.4 billion in advertising revenue for 2026 and roughly $100 billion by 2030, based on earlier reports from The Information and Axios.

Those forecasts are considerably higher than current estimates for the wider chatbot advertising market. EMARKETER expects US chatbot advertising revenue to remain below $1 billion in 2026.

Nate Elliott, an analyst at EMARKETER, told Business Insider that OpenAI would need a much larger user base, stronger consumer adoption, aggressive pricing, and a high volume of advertisements to meet its projection.

ChatGPT has about one billion monthly users, giving OpenAI a large potential audience. The company has so far limited the number of advertisements shown as it tests how advertising affects the product experience.

Early pilot campaigns involved major advertisers committing hundreds of thousands of dollars. OpenAI advertising chief Dave Dugan said the company restricted inventory during the initial rollout to prioritise the user experience.

OpenAI’s restricted inventory limits campaign scale. Advertisers interviewed by Business Insider also said they receive less campaign data than they do from established platforms.

The company introduced a self-service tool in May, allowing advertisers to create and manage campaigns directly. Geographic targeting within the US followed in July, replacing the earlier requirement to target the country as a whole.

OpenAI’s Ads Manager Beta allows businesses to set budgets and bids, upload advertisements, launch campaigns, and review performance. Its current tools also support conversion measurement and cost-per-click bidding.

Targeting conversations rather than keywords

OpenAI says ChatGPT considers the context and intent of the current conversation when selecting an advertisement. The system also reviews the advertisement’s copy, landing page, and information supplied by the advertiser.

Advertisers can submit “context hints” describing the topics, needs, or situations in which a product might be relevant. These hints guide the matching system but do not operate as exact targeting rules.

Advertisers can control budgets, bids, pacing, and creative materials. OpenAI retains control over which eligible advertisement is delivered and the conversation in which it appears.

That differs from keyword-based search advertising. Google advertisers can review the search terms that triggered their advertisements, change keyword bids, and use negative keywords to block irrelevant queries.

OpenAI says users often research categories, compare options, and make decisions within a conversation, allowing relevance to be assessed across several exchanges rather than one search term.

Brands therefore have to describe the conversational situations in which a product is relevant rather than depend only on a fixed list of keywords. OpenAI’s system then decides whether the wider discussion matches those instructions.

Speechify, an AI text-to-speech company, has advertised through ChatGPT since the early pilot programme. Chief business officer Rohan Pavuluri said the company’s advertisements appeared alongside comparisons of text-to-speech tools, generating new customer activity.

Pavuluri said Speechify’s advertisements appeared below text-to-speech comparisons, placing them alongside conversations about a relevant product category.

Kevin Roy, CEO of agency GreenBanana SEO, said advertisers are interested in reaching ChatGPT users but need to structure campaigns around user prompts rather than keywords. Advertisers also receive less campaign data than they would from established platforms.

Roy said ChatGPT provides substantially fewer campaign controls than Google Ads. The difference limits how advertisers can adjust targeting, placements, bidding, and other campaign settings.

Marketing startup Launch10 recorded a lower click-through rate on ChatGPT than on Google Ads during a trial campaign, according to CEO Greg Hockenbrocht. He had expected the platform to include more automated features for generating advertising copy and testing different versions.

Hockenbrocht said performance improved during a month-long campaign, with impressions and clicks increasing as the campaign progressed. He also pointed to the introduction of location targeting as evidence that the platform was adding capabilities.

Speechify reported new business from advertisements shown under relevant product comparisons. Launch10 reported a lower click-through rate than Google Ads and said it wanted more creative-testing tools.

The two campaigns provide different results from the same emerging format. Speechify benefited from placement beside a relevant product discussion, while Launch10’s experience showed that contextual matching does not remove the need for testing, reporting, and optimisation controls.

OpenAI said it is collecting feedback from advertisers and continuing to develop the service. The company did not respond to EMARKETER’s assessment of its revenue expectations.

Measurement without conversation-level visibility

OpenAI initially sold ChatGPT advertising on a cost-per-thousand-impressions basis before introducing cost-per-click buying. The company has also added conversion measurement and broader performance reporting.

OpenAI’s public documentation describes cost-per-click buying, conversion measurement, and aggregated campaign reporting, but it does not describe reporting that attributes an advertisement to an individual prompt or stage of a conversation.

Google’s reporting tools allow advertisers to review performance by campaign, advertisement, ad group, keyword, and search term. Marketers can use those results to adjust bids, remove weak terms, or exclude irrelevant searches.

OpenAI’s public materials do not describe a prompt-level reporting system comparable with the keyword and search-term reports available through Google Ads.

Advertisers interviewed by Business Insider said they receive less campaign data than they do on established platforms. OpenAI says advertisers do not receive the underlying conversations or users’ personal details.

A conversational advertising model may therefore require reporting organised around broader topics, needs, or decision stages rather than individual keywords. OpenAI has not publicly detailed whether its reporting will develop in that direction.

The distinction is also relevant to creative testing. Hockenbrocht said he expected more tools for generating ad copy and comparing different versions when Launch10 tested the platform.

OpenAI has added campaign performance and conversion reporting, while advertisers continue to ask for more detailed testing and optimisation tools.

The current system can report whether an advertisement generated an impression, click, or conversion. Its public documentation does not explain how advertisers can identify which conversational context contributed to that result.

Privacy limits the data advertisers receive

OpenAI says conversations remain private from advertisers and that it does not sell user data. Advertisers only receive information that users choose to share with them directly through an advertisement.

OpenAI says its measurement tools allow businesses to assess campaign performance without sharing conversations or personal details with advertisers.

This limits the level of conversational detail that can appear in advertiser reports. OpenAI must provide enough information for campaign evaluation without disclosing the underlying thread used to assess relevance.

OpenAI separates contextual targeting from ad personalisation. When personalisation is enabled, the system can use information from previous chats and earlier advertising interactions alongside the current conversation.

Users who disable personalisation can still receive advertisements based on the subject of the current thread. Previous chats and advertising activity are not used in that case.

The distinction means contextual targeting can continue without using a person’s wider ChatGPT history. It also gives advertisers no direct access to the individual exchanges behind the placement.

OpenAI also allows users to clear advertising data, dismiss advertisements, and review why an advertisement appeared. The company says advertisements are displayed separately from ChatGPT’s answers and do not influence the model’s responses.

OpenAI says a paid placement does not constitute an endorsement or recommendation from the company.

OpenAI’s policies also restrict advertisements from appearing around sensitive or unsuitable conversations. These include discussions involving personal health, emotional vulnerability, self-harm, political content, illegal activity, and other high-risk subjects.

The system must therefore assess both commercial relevance and whether the surrounding conversation is suitable for advertising.

A developing channel with large revenue targets

ChatGPT advertising remains a developing channel. OpenAI’s self-service campaign manager is still in beta, while the company is gradually expanding its geographic reach and campaign capabilities.

Advertisers currently have access to conversational context, bidding, conversion measurement, location targeting, and self-service campaign management. They continue to receive fewer testing, targeting, and reporting controls than are available through Google and Meta.

The platform is still being tested by early advertisers rather than operating with the targeting, measurement, and optimisation infrastructure available through mature advertising networks. The source material does not establish that advertisers are moving a substantial share of their media budgets to ChatGPT.

Its current role is therefore better measured through campaign tests, advertiser feedback, and product development than through direct comparisons with the scale of Google or Meta.

OpenAI said it is gradually opening Ads Manager to more businesses while continuing to test and refine the platform. The company also expects to add further bidding and optimisation options over time.

(Photo by Levart_Photographer)

See also: Google leads AI discovery as ChatGPT dominates LLM referrals

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

Marketing Tech News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.