A product carousel is a ChatGPT Ads format that displays several products from a single retailer side by side in one sponsored placement at the bottom of a conversation. The original single-product ChatGPT ad promotes one item with an image, description, and link, while the carousel lets shoppers compare multiple options without leaving the chat.
For ecommerce marketers, this moves the lever of visibility from ad creative to catalog data, because OpenAI decides when a carousel appears based on your product feed. If your feed carries weak titles, missing prices, or poor images, your products can lose the placement to a retailer whose data is cleaner.
Product carousels sit in the same slot ChatGPT uses for a single sponsored result, but they draw multiple items from one retailer's catalog and present them as a scrollable row of cards. OpenAI assembles the unit automatically from an uploaded product feed instead of a hand-built ad.
Each card typically carries a product image, a short piece of text, a price, and a link to the advertiser's site. The feed behind it uses the same file structure as a Google Shopping feed, so the titles, descriptions, prices, images, and availability you maintain there become the raw material OpenAI chooses from. The platform decides whether a given query triggers a single product or a full carousel, and the advertiser does not make that call.
This places a product carousel closer to a shopping placement than to a conversational text ad, and it serves only free and lower-cost ChatGPT users in eligible markets. AirOps helps ecommerce teams keep the catalog and content signals behind these placements accurate and on-brand as AI shopping surfaces shift.
Resources: See how to structure feeds, creative, and policy for ChatGPT Ads placements
A product carousel runs on retailer catalog data and OpenAI's automated ad-serving system, moving from feed to placement to measurement.
Upload the feed: The retailer sends its product catalog to OpenAI's ad platform, including titles, prices, images, and availability for each item.
Match the query: When a user asks a shopping-related question, OpenAI's systems scan the conversation and identify products in the feed that fit the intent.
Choose the format: The platform decides whether that query is better served by a single product card or a multi-product carousel, and the advertiser cannot force either one.
Render the cards: ChatGPT displays the selected products side by side at the bottom of the conversation, clearly labeled as an ad, with each card linking to the advertiser's site.
Measure the click: Dynamic URL parameters and third-party attribution pass the click back to the retailer's analytics so conversions can be tracked.
The output tells you which products OpenAI found relevant enough to surface and how shoppers engaged with them. It does not tell you when a carousel will appear in place of a single ad, because that decision stays inside OpenAI's systems.
Product carousels change where ecommerce spend buys visibility inside ChatGPT. The deciding factor is the quality of your product feed, since OpenAI selects which items to surface directly from that catalog data. Your ad spend now follows the health of the feed, so merchandising and data hygiene become budget decisions.
Feed quality decides placement: Because OpenAI builds the carousel from your catalog, weak titles, outdated prices, or low-resolution images can push your products out of the slot entirely and hand the impression to a cleaner feed.
Comparison happens in the ad: Shoppers weigh several products together before they click, so a card that lacks a compelling price or image loses the comparison even when it is shown alongside your other items.
A new shopping surface opens before Q4: OpenAI rolled the format out ahead of the holiday season, giving retailers a conversational discovery channel that sits alongside Google Shopping and paid search placements.
Growth marketers use product carousels to put their best-converting SKUs in front of shoppers who are actively asking ChatGPT for product recommendations.
SEO managers use product carousels to extend catalog visibility into AI chat surfaces that their organic shopping listings cannot reach.
Demand gen leads use product carousels to test whether conversational ad placements drive qualified ecommerce traffic before committing larger Q4 budgets.
Your product feed is the structured catalog file OpenAI reads to assemble every carousel, which means the completeness and accuracy of its titles, descriptions, prices, images, and availability fields directly determine which of your items are even eligible to appear and how appealing they look beside competing options.
OpenAI's automated systems decide when a multi-product carousel replaces a single-product ad for a given conversation, so marketers cannot select the format manually and instead shape the result by strengthening feed inputs, product relevance, and the clarity of the data OpenAI draws on.
Each carousel currently displays items from one retailer at a time, so your products are presented together as a curated set from your own catalog, and the assortment a shopper sees depends entirely on how broad and well-structured your uploaded feed happens to be.
Reach shoppers inside ChatGPT at the moment they ask for product recommendations.
Showcase multiple SKUs in a single placement instead of one item per ad.
Compete on catalog data you already maintain for Google Shopping.
Scale campaigns from a product feed without building ads one by one.
Open a new conversational shopping channel ahead of the Q4 holiday season.
Measure clicks and conversions through dynamic URL parameters and third-party attribution.
Keep product titles specific and front-load the attributes shoppers search for, because OpenAI builds the card directly from that text.
Maintain accurate prices and availability in the feed, since a stale price or an out-of-stock item wastes the placement.
Use high-resolution, clean product images, as the visual carries most of the comparison in a scrollable row of cards.
Segment your feed so your strongest-margin, best-converting SKUs stay eligible, because you cannot choose which items the carousel surfaces.
Add dynamic URL parameters for attribution, so you can see which carousel clicks actually convert into revenue.
Mirror the feed hygiene you already apply to Google Shopping, since the two formats share the same underlying file structure.
Avoid treating the product feed as a set-and-forget upload. The most common mistake competent teams make is loading a catalog once and leaving titles, prices, and images to drift, which quietly erodes eligibility as OpenAI's systems favor fresher, more complete data and move the placement to a retailer who keeps their catalog current.
AirOps: Shows where your ads and competitors' ads appear across the prompts you track in ChatGPT, so you can defend brand-relevant shopping queries and keep the catalog signals behind your carousels on-brand.
OpenAI ChatGPT Ads Manager: The platform where retailers upload product feeds, set targeting and budgets, and run the campaigns that power carousel placements.
Google Merchant Center: Manages the Shopping-format product feed whose titles, prices, and images can be reused as the data source for ChatGPT carousels.
Audit your feed: This week, open your existing product catalog and check titles, prices, images, and availability for the SKUs you most want surfaced. No budget or approval is needed to start here.
Fix data gaps: Correct missing attributes, low-resolution images, and stale prices so your highest-priority products are both eligible and attractive in a side-by-side row.
Set up a ChatGPT Ads account: Create or access your OpenAI ad platform account and confirm your business meets the eligibility requirements for product feed campaigns.
Upload the product feed: Connect or import your catalog, reusing your Google Shopping feed structure where possible to save setup time and keep one source of truth.
Add attribution and measure: Apply dynamic URL parameters and connect your analytics so you can track which carousel clicks convert, then refine the feed based on what performs.
A product carousel is a ChatGPT Ads format that shows several products from one retailer side by side at the bottom of a conversation.
The unit is assembled automatically from an uploaded product feed built in the same structure as a Google Shopping catalog.
OpenAI decides when a carousel appears and the advertiser cannot select the format, which is the main constraint of the channel.
Weak feed data such as poor titles, stale prices, or low-quality images is the main risk, because it can quietly cost you the placement.
Feed quality is where the leverage sits, since improving catalog inputs is the clearest way to influence which products get surfaced.
A product carousel shows several items from one retailer in a scrollable row, while a single-product ChatGPT ad promotes just one item with an image, short description, and link. Both appear in the same slot at the bottom of a conversation and are clearly labeled as ads, so the difference the shopper notices is the number of options and the ability to compare them in place. The deeper difference is how each is built: a single-product ad can be a hand-built unit, while a carousel is assembled automatically from your product feed. That shifts your work from writing individual ad copy to keeping an accurate, well-structured catalog. It also changes how you win the slot, because the carousel pulls many SKUs from the same feed and presents them together, so the strength of your overall catalog matters as much as any one product listing.
ChatGPT shows a product carousel when OpenAI's ad system judges that a conversation is better served by several comparable products than by one. You do not set this yourself. The platform reads the intent of the current conversation, checks which items in your feed are relevant, and then chooses between a single card and a multi-product row. In practice that means carousels tend to surface on shopping-style questions where a shopper is weighing options, such as asking for recommendations within a category. Because the choice sits inside OpenAI's systems, the same feed can produce a single ad in one conversation and a carousel in another. The most reliable way to be eligible for either is to keep a broad, clean feed, so that whenever OpenAI does decide a carousel fits, your products have the data quality to be among the ones it selects.
Product carousel placements vary because OpenAI selects products conversation by conversation, matching each query's intent against your feed in real time. The same prompt phrased two ways, or asked by two different users, can surface different items or shift between a single ad and a full carousel. Several inputs feed that decision: the wording and context of the conversation, the relevance and completeness of your catalog data, and OpenAI's own serving logic, which the advertiser does not see or control. Eligibility and competition also move, so a product that appears today may be edged out tomorrow by a listing with a sharper title, a better price, or a stronger image. This variability is normal for an automated, intent-driven channel. The practical response is to monitor which products surface over time and to keep feed quality high, since no single SKU holds a fixed placement.
No, you cannot directly pick which items land in a product carousel, because OpenAI's system selects them from your feed based on the conversation. What you can control are the inputs that make selection more likely to go your way. Start with the feed itself: include the SKUs you want surfaced, write accurate and specific titles, maintain current prices and availability, and supply clean, high-resolution images. Segment or prioritize your catalog so your strongest-margin and best-converting products are present and well described, since the platform can only choose from what you upload. You also influence outcomes through relevance, making sure the products and their data clearly match the kinds of questions your buyers ask. Think of your job as shaping a high-quality pool of eligible products and letting OpenAI match them to conversations, instead of trying to hand-place a specific item in a specific slot.
A good product carousel feed is complete, accurate, and current, so OpenAI has strong data to choose from and shoppers see compelling cards. Titles should be specific and lead with the attributes people actually search, descriptions should be clear and concise, and prices and availability should reflect real-time stock so the placement is never wasted on an item a shopper cannot buy. Images matter heavily in a side-by-side row, so use clean, high-resolution visuals that show the product clearly. Breadth helps too: a feed that covers your relevant catalog gives the system more eligible items to assemble into a carousel. Because the format reuses the Google Shopping file structure, the feed hygiene you already practice there transfers directly. The clearest sign of a healthy feed is consistency over time, with regular updates that keep every field trustworthy as your catalog, pricing, and inventory change.