OpenAI unveils ad tech to compete for marketing budgets
▼ Summary
– OpenAI is adding product carousels to ChatGPT ads, using product feeds to display multiple items from a single retailer at the bottom of conversations.
– OpenAI decides whether to show a single product or a carousel, giving advertisers no control over the format.
– A new AppsFlyer integration lets advertisers attribute app installs, purchases, and subscriptions to ChatGPT ads, with about 40 brands like Grubhub testing it.
– The integration provides conversion evidence but does not measure incrementality or address reporting and optimization questions.
– OpenAI has built the basic infrastructure for performance advertising, but must prove it can deliver conversions at scale to earn sustained advertiser budgets.
OpenAI is rolling out new advertising capabilities for ChatGPT, including product carousels and a partnership with AppsFlyer for conversion tracking. These moves position the company to compete for performance marketing budgets by giving advertisers both shoppable ad formats and measurable outcomes.
The new carousel format displays multiple products within a single ad placed at the bottom of a ChatGPT conversation. Previously, ads on the platform were largely limited to featuring a single product. The shift gives brands a more visual, catalog-driven way to reach users mid-conversation.
This format builds on the product feeds feature OpenAI introduced roughly three months ago. Retailers submit their catalogs, and OpenAI uses that data to generate either single-product ads or carousels. Importantly, advertisers do not choose which format appears. OpenAI decides whether to show one item or a rotating selection, which means the platform retains control over product presentation. Currently, each carousel pulls products from a single retailer only.
On the measurement side, OpenAI has teamed up with AppsFlyer to link ChatGPT ads to app installs, in-app purchases, and subscription events, as reported by Adweek. Around 40 brands, including Grubhub, are already testing the integration. App marketers can now fold ChatGPT into their existing acquisition dashboards and compare its attributed conversions against other paid channels.
It is worth clarifying that this integration does not measure incrementality, nor does it answer deeper questions about reporting or optimization. What it does provide is concrete evidence of whether ChatGPT ads lead to conversions, which is a necessary first step for brands weighing budget allocation.
Taken together, these updates signal that OpenAI is assembling the core components of a performance ad system. In a short span, the company has introduced automated product feeds, dynamic ad presentation, and third-party attribution. Retailers supply the product data, OpenAI manages how products are displayed, and advertisers can now track results. That combination gives ChatGPT the foundational infrastructure needed for performance advertising.
Looking ahead, OpenAI is reportedly offering advertisers more guidance on feed-based campaigns in preparation for Q4 and the holiday shopping season. This period will serve as a real test of whether improved merchandising and measurement can attract larger ad budgets.
Still, advertiser control remains a sticking point. Since OpenAI determines whether a single product or carousel is shown, marketers will want clarity on how those decisions are made, how individual products perform, and what level of control they can expect going forward.
Scale is another open question. Better formats and attribution make ChatGPT easier to test, but they do not yet prove the platform can deliver enough conversions at competitive costs to justify sustained investment. OpenAI has built much of the machinery required for performance advertising. Now the challenge is demonstrating it can produce results at scale.
(Source: MarTech)
