OpenAI aims to earn $11 billion from ChatGPT advertising by 2027, but the pilot CPM is only sold at $15.

ME News Report, April 16 (UTC+8), according to Beating Monitoring, OpenAI is about to introduce a pay-per-click (CPC) model for ChatGPT ads and is exploring conversion-oriented ad formats (driving purchases or app downloads), but the latter has no clear timetable yet.
The current state of advertising business shows a significant gap between ambitions and reality. OpenAI began displaying ads to logged-in U.S. users on the free and lowest-priced paid versions of ChatGPT in February this year, initially targeting a CPM (cost per thousand impressions) as high as $60, comparable to premium advertising slots like streaming TV.
However, some advertisers’ actual achieved CPMs are only between $15 and $25, possibly reflecting a shortage of bidders for ad slots. The pilot period has also been extended: ad frequency was insufficient, and early advertisers’ budgets had not been fully spent by the end of March.
At the end of March, OpenAI stated in a blog that the six-week pilot exceeded $100 million in ARR, but did not specify the calculation method.
In Q1, OpenAI told investors that it expects advertising revenue to reach $2.4 billion this year and $11 billion by 2027. If the latter target is achieved, its ad scale will far surpass Snap and Pinterest.
The core dissatisfaction among advertisers is the lack of effect tracking. Meta and Google provide detailed audience profiles and conversion attribution, while OpenAI currently only offers aggregated data (impressions, clicks, spend).
Ben Kahan, head of programmatic business at marketing agency Brainlabs, said: “Many clients with budgets wanting to try new channels are watching cautiously because they feel they can’t get the desired effect data.”
OpenAI is accelerating its learning curve. In recent weeks, it has opened a self-service management backend to some advertisers, who previously relied on spreadsheets and phone communication; it has partnered with ad tech company Criteo for resale; new signed advertisers’ minimum monthly commitments have been lowered to $30k–$50k, with early prepayments as high as $200k.
However, targeting capabilities remain a weakness, as advertisers can only provide broad keywords as guidance, unable to target specific prompts or user types precisely.
Ad formats are also limited to short titles with small images, which Jellyfish Chief Solutions Officer Jai Amin described as “quite old-fashioned.”
The real challenge for OpenAI in advertising is not technology but managing advertisers’ expectations.
Traditional digital advertising has evolved over twenty years into a complete attribution system from exposure to conversion, with advertisers accustomed to pay-for-performance and precise targeting.
ChatGPT’s conversational interaction naturally makes this logic difficult to apply: each response is dynamically generated, with no fixed page to anchor an ad slot, and it cannot bid on keywords like search ads.
Shifting from CPM to CPC and conversion-based billing is the right direction, but to truly compete with Meta and Google, OpenAI needs to address not only billing models but also the entire advertising measurement and targeting infrastructure.
(Source: BlockBeats)