If you run a retail brand and you’ve been watching the ChatGPT advertising space, this week’s news is worth paying attention to. OpenAI just rolled out a product feed ads feature inside its Ads Manager beta, and it changes how retailers can advertise on the platform in a pretty significant way.
Up until now, retailers had to build campaigns one product at a time to run ads on ChatGPT. But for a brand with thousands of items in its catalog, that’s just not practical. This new product feature changes that entirely.
What is the New Feature from OpenAI
OpenAI introduced a product feed option within the ChatGPT Ads Manager where retailers could input their product catalogs and automatically generate ads directly from their catalog items without having to create individual ads for all items manually.
As OpenAI confirms in its Help Center, the platform can handle up to one million SKUs per advertiser but requires a minimum of 100 SKUs to start with. For a retailer carrying tens of thousands of products, this is the feature that finally makes running ChatGPT ads at scale actually workable.
Interestingly enough, OpenAI even stated that feed ads had proven themselves to be some of the most effective in their Ads Beta test run. This information came to light after the founder of JXT Group, Menachem Ani, posted an email he got from OpenAI announcing the introduction of the new functionality.
How the Ad Format Actually Works
This works more like Google Shopping and Performance Max than Amazon Sponsored Products. Advertisers don’t pick keywords or set up manual targeting. The system builds ads automatically from product names, images, and catalog attributes, then serves them below ChatGPT responses.
So, if a user engages in a conversation with ChatGPT and asks for advice about running shoes priced less than $100, a retailer whose products include such shoes can display its product in the form of an ad below ChatGPT’s response at the very moment when the customer needs it.
This is where the added value of the product feed advertising lies. People in those conversations are already close to buying something, so the purchase intent is naturally high.
How This Compares to Google and Meta
This move isn’t entirely new ground for the advertising industry. Google has had dynamic product feeds in Google Shopping for years. Meta runs something similar with its catalog-based dynamic ads, where retailers upload their inventory and Facebook or Instagram automatically serves relevant products to users based on their browsing behavior and interests.
What OpenAI is doing is taking that same proven model and bringing it to a conversational AI platform. The core idea, matching a retailer’s inventory to user intent automatically and at scale, is the same. The context is just different.
With Google and Meta, the targeting relies on search queries or behavioral signals from past browsing. With ChatGPT, the intent signal is live and conversational. A user literally types out what they need in plain language, and the system can match that in real time to a relevant product in a retailer’s catalog.
That’s potentially a much cleaner signal than a search keyword or a retargeting pixel.
The Bigger Picture for OpenAI’s Ad Business
OpenAI’s advertising offering was rolled out in February 2026, and it didn’t take long for them to scale quickly from the limited beta program with heavy minimum spends to a self-service platform. This release of the product feed feature marks yet another step forward on this path to scalability and reach for retailers on the platform.
Industry analysts project that AI-powered search advertising in the U.S. will grow from around $1.1 billion in 2025 to roughly $26 billion by 2029, and OpenAI is clearly positioning itself to grab a meaningful share of that.
Furthermore, the company announced the addition of conversion tracking functionality to its Ads Manager Beta and released cost-per-action bidding in early June 2026.
What Retailers Need to Think About
One thing worth noting is that a retailer’s product feed quality matters a lot here. The ads the system generates are only as good as the catalog data it has to work with. Clean product titles, good images, accurate pricing, and well-structured attributes are going to directly affect whether the auto-generated ads are compelling or not.
According to data shared at a May 2026 panel, roughly 95% of merchants are already seeing AI agent traffic on their sites, but only about 20% have machine-readable catalogs set up properly. That gap is now both an organic visibility issue and a paid advertising eligibility issue at the same time.
Retailers who already send a product feed to Google Shopping are in the best position to jump into this quickly, since ChatGPT Ads Manager uses the same structured catalog format. It doesn’t require starting from scratch.
What’s Coming Next
Retail advertisers in the beta can review the feed requirements and start building campaigns directly from their uploaded catalogs. If you’ve been on the fence about testing ChatGPT ads at scale, now is a good time to take a serious look.
The advertising infrastructure within ChatGPT is still relatively new. But things are moving quickly, and the product feed launch makes it clear that OpenAI is serious about making ChatGPT a real retail advertising channel.
Frequently Asked Questions
Q 1: What are product feed ads in ChatGPT?
Product feed ads let retail advertisers upload their full product catalog to OpenAI’s Ads Manager and automatically generate sponsored ads from those catalog items. Instead of building each ad manually, the system creates placements using product names, images, and attributes from the feed.
Q 2: How many products can a retailer upload to ChatGPT Ads Manager?
OpenAI’s system supports up to one million SKUs per advertiser. Retailers start by connecting a 100-product sample before opening up access to their full catalog.
Q 3: Do advertisers need to pick keywords for ChatGPT product feed ads?
No. The platform works more like Google Shopping or Performance Max, where the system handles targeting automatically based on conversational intent. Advertisers do not set up manual keywords or match types.
Q 4: How is ChatGPT advertising different from Google Shopping?
With Google Shopping, targeting relies on search queries. With ChatGPT product ads, the intent signal comes from live conversations, meaning users are actively describing what they need in real time, which can be a much more direct signal of purchase intent.
Q 5: Who first reported the ChatGPT product feed ads launch?
The update was first shared by Menachem Ani, founder of JXT Group, who posted the email he received from OpenAI about the new feature on X (formerly Twitter).
Q 6: Are ChatGPT product feed ads available to all retailers right now?
The feature is currently in Ads Manager beta. Retail advertisers in the beta program can review feed requirements and begin creating campaigns from their uploaded product catalogs.
Q 7: What makes feed-based ads perform well on ChatGPT?
OpenAI has said that feed-based ads have been among the strongest-performing formats in its Ads Beta program so far. The likely reason is that product ads shown during purchase-focused conversations reach users who are already in a buying mindset.
Q 8: Does a retailer need a special format for their product feed?
ChatGPT Ads Manager uses the same structured catalog format that retailers already send to Google Shopping, so brands already running Google product ads can use an existing feed without rebuilding it from scratch.
Q 9: What measurement tools does ChatGPT Ads Manager offer?
Conversion measurement is live inside Ads Manager Beta, and cost-per-action bidding began rolling out in early June 2026 for accounts with at least one conversion event already flowing. Post-view attribution is also on OpenAI’s public roadmap.
Q 10: Why is this launch significant for the retail advertising industry?
It brings a proven format, dynamic catalog-based advertising, to a conversational AI platform for the first time at real scale. It also signals that OpenAI is building ChatGPT into a serious paid advertising channel for retailers, not just a content tool.
