ChatGPT Ads: From Attention Economy to Intelligence Economy
OpenAI’s ChatGPT advertising platform is a conversational ad format that appears in a sponsored box beneath organic chatbot answers, positioned as mid-funnel marketing inventory where users research and compare options rather than passively browse or search. This is not an incremental banner play; it is an attempt to reframe the ad market from an attention economy to what OpenAI calls an “intelligence economy,” where context-rich conversations replace bare keywords as the primary signal of intent. At Cannes Lions, OpenAI used its debut to pitch this emerging OpenAI ad business as a new way for brands to meet people who arrive with a “job to be done,” not a casual scroll. That framing is the real story: the company wants marketers to see ChatGPT as the moment when decision-making is in progress, and where ads can feel like help rather than interruption.
Mid-Funnel Positioning and Falling AI Advertising Dismiss Rates
OpenAI is explicit: ChatGPT ads are “decision-layer inventory, effectively mid-funnel,” reaching people as they research topics in depth and compare products. Around one-fifth of queries show direct commercial intent in categories like travel, retail, health and beauty, and financial services, which makes this inventory look more like intent-rich search than social feed impressions. The ads appear in a clearly labelled sponsored box below organic responses, are shown only to free and Go tier users, and do not influence the chatbot’s answers. More telling than the design is early behaviour: OpenAI tracks “cross-out rates,” when users dismiss an ad, and says this AI advertising dismiss rate has fallen by 50% since launch. That is the quotable proof point OpenAI needed—evidence that conversational ads can scale without users angrily shutting them off.
From Experiment to Platform: Adoption, Bidding and Measurement
This is no longer a quiet experiment. ChatGPT ads have expanded to seven markets since February, with additional territories already flagged. More than 2,000 brands are buying inventory through an ad-tech partner, signaling genuine demand for the ChatGPT advertising platform rather than curiosity clicks. OpenAI has rolled out a self-service Ads Manager, moved beyond cost-per-thousand impressions to include cost-per-click bidding, and reports that CPC has quickly become popular with performance-focused marketers. According to the Financial Times, OpenAI spent $34 billion last year, with costs outpacing revenue, which makes this ad push an economic necessity as much as an innovation story. On measurement, the LiveRamp partnership is key: advertisers can use LiveRamp’s Conversions API Hub to track performance beyond unreliable browser-based signals, while integration with Criteo, Adobe, StackAdapt and major agencies keeps ChatGPT ads plugged into familiar workflows.
Why OpenAI Is Targeting Google and Meta’s Turf
OpenAI is not shy about its ambitions. It used Cannes to sell ChatGPT as “something brand new to the market” and a genuine alternative channel to reach consumers, directly challenging search and social incumbents. Ads appear when people actively seek products or services, mirroring the logic of search advertising while leaning on richer conversational context. Unlike keyword auctions, ChatGPT sees full problem statements, comparisons and constraints, which OpenAI argues can reveal emerging needs even before explicit shopping intent appears. The strategic bet is clear: if mid-funnel intent can be captured in conversation, OpenAI can carve out a slice of the performance budgets that have long defaulted to Google and Meta. Yet industry executives remain skeptical that this format alone can deliver search-scale revenue quickly, and expect OpenAI will need more ad products and formats to unlock major brand spend. For now, conversation is the differentiator—and the risk.
A New Kind of Inventory, With Old Measurement Questions
The most interesting part of OpenAI’s pitch is not the creative, but the rules. Sponsored content is kept separate from organic answers, conversation data is not shared with advertisers, and users can clear ChatGPT history, all framed as trust principles meant to avoid the backlash that followed other scaled ad platforms. That constraint-heavy approach creates a tension: marketers want precise targeting and multi-touch attribution, while OpenAI wants to protect the integrity of the chat experience. For now, the company is urging brands to design “answer-plus-next-step” creative that responds directly to the user’s question, treating query context as the core signal rather than demographic profiling. This makes ChatGPT mid-funnel marketing inventory feel less like a blunt performance tool and more like consultative selling. If OpenAI can keep AI advertising dismiss rates low while proving incremental lift via partners like LiveRamp, it will have built not only a new revenue stream, but a new kind of ad channel.
