Earlier this year, OpenAI launched its advertising business on ChatGPT, initially partnering with well-known brands with substantial advertising budgets such as Adobe, Ford Motor Company, and Target. Now, the company is expanding its advertising operations, introducing more services tailored for small and medium-sized enterprise marketers, positioning itself to compete directly with advertising giant Meta.
A person familiar with OpenAI's management revealed that company executives have informed several external advertising technology firms that a key current objective is to attract local small business clients, such as car washes and dry cleaners. Another advertising agency executive noted that OpenAI has also rolled out a series of customized features for small and medium-sized advertisers, including conversion-focused ads designed to prompt users to complete specific actions, such as making a purchase, booking a service, or filling out a contact form.
This move signifies OpenAI's efforts to catch up with the established capabilities of veteran advertising giants like Meta and Google. According to the aforementioned advertising agency executive, OpenAI has informed advertisers that, starting this week, they can test these conversion-based ads—where advertisers only pay when an actual conversion occurs. To run such ads, clients must install OpenAI's advertising pixel code on their websites to track user actions following an ad click.
While advertising pixels are a common industry technology, they are not entirely accurate, as browsers or ad-blocking plugins can sometimes block the code. To address this, OpenAI also recommends that advertisers integrate their own systems with its recently launched application programming interface (API). Through this interface, advertisers can transmit data on user behavior on their websites, helping OpenAI verify actual ad conversion results.
Due to restrictions from user privacy regulations, the effectiveness of traditional technologies like advertising pixels has been diminishing, leading major advertising platforms to introduce supporting API interfaces in recent years.
In addition, OpenAI is engaging with various external advertising technology companies to explore collaborations, focusing on addressing the contextual display of ads in conversational settings to alleviate advertiser concerns about new ad formats. According to sources, OpenAI's advertising business leads overseeing these partnerships include Dave Dugan, who joined in late March to head advertising sales, along with Benji Shomair, Vijay Raj, and Asad Awan. These executives previously held long-term positions at Meta.
The recent series of new features build upon OpenAI's earlier updated cost-per-click advertising model. The platform initially used a cost-per-impression model, where advertisers paid regardless of user interaction, which had limited appeal for most advertisers.
Another advertising agency executive mentioned that OpenAI also plans to integrate product catalogs with ad delivery to serve e-commerce merchants. This feature may leverage technology from the previously discontinued in-chat checkout project, using merchant product libraries to serve ads highly relevant to ChatGPT conversation content. The launch timeline for this feature remains undetermined.
Overall, these adjustments aim to address a common complaint from advertisers: ChatGPT ads lacked basic data monitoring tools, making it difficult to assess campaign effectiveness. This is particularly crucial for small and medium-sized advertisers, who prioritize quantifiable conversion outcomes over mere brand exposure. Small and medium-sized advertisers are a significant source of advertising revenue for Meta and Google.
An OpenAI spokesperson stated that the company will continue to introduce more tools to help advertisers optimize their campaigns and focus on key conversion goals.
**Lowering the Barrier to Entry**
Advertising is a key revenue-generating initiative for OpenAI. The platform currently has 900 million weekly active users, the majority of whom use the free version. The company disclosed to investors that advertising revenue is projected to reach $2.4 billion this year; by 2027, this revenue could quadruple; and by 2030, advertising revenue is expected to hit $102 billion, accounting for over 35% of the company's total revenue.
OpenAI is collaborating with investment banks to prepare for an initial public offering, potentially launching as early as this year. Rapid growth in advertising will enhance its appeal to capital markets. However, while Meta and Google have built vast advertising ecosystems supported by a massive base of small and medium-sized advertisers, the growth rate of OpenAI's advertising business remains uncertain.
When ChatGPT ads officially launched in early February, OpenAI primarily worked through large advertising agencies to partner with top brands, with a minimum initial campaign threshold of $200,000. Within six weeks, its advertising business achieved an annualized revenue run rate of $100 million.
In the initial phase, OpenAI relied on its in-house sales team to onboard the first wave of advertising clients. It quickly expanded its partnership ecosystem, collaborating with external advertising platforms like Criteo and StackAdapt, which assist their clients in running ChatGPT ads. These partners also provide capabilities OpenAI currently lacks, such as precise user matching technology.
Now, with its services open to small and medium-sized advertisers, OpenAI has also eliminated the minimum prepaid spending requirement. Several advertising agencies have revealed that a self-service ad platform is now available in the U.S. for businesses of all sizes, with no minimum spend. Most external advertising platforms partnering with OpenAI have similarly removed campaign thresholds.
Some advertising agencies indicate a preference for using third-party platforms over direct engagement with OpenAI, as third parties can offer additional services.
Deanna Cullen, Vice President of Wipro's media buying division, stated, "Through these partnerships, both parties can engage in deep collaboration across data integration, e-commerce scenarios, and creative advertising, offering vast development potential." The company serves clients including Volvo and Alizia.
Currently, there is no public information on the revenue-sharing details between OpenAI and its partners, including whether the platform pays referral fees to partner channels or how much service fees advertisers pay to intermediaries. Even if OpenAI does not directly subsidize partners, the intermediary fees paid by advertisers could reduce the revenue ultimately flowing to OpenAI.
Leveraging external partners to expand advertising operations is common in the tech industry. Shortly after its 2012 IPO, Facebook launched an ad exchange platform, allowing third-party agencies to assist clients with ad placements. At the time, Facebook's own advertising tools were still underdeveloped, and this move effectively expanded its advertiser base.
Cullen and other advertising agency executives noted that currently, most clients still treat ChatGPT advertising as experimental, not allocating funds from their fixed advertising budgets. Data shows that initial ad click-through rates have been impressive, but such emerging ad networks often experience a decline in performance as ad volume increases and user novelty wanes.
As for features like product catalog-linked ads, implementation challenges remain unclear. These features require standardized data on pricing, inventory, and product descriptions to serve precise ads matching conversation content. This process is complex: the in-chat shopping project that OpenAI halted in March faced slow progress primarily due to disorganized merchant product data.
Comments