ChatGPT product recommendations: how products get chosen, and how to make yours eligible

ChatGPT product recommendations are an organic shopping surface where the model suggests products, merchants, and prices based on a user’s query, not on paid placement. As of October 2026, OpenAI documents 2 eligibility paths for merchants and a ranking system that uses product and merchant data rather than bids. For decisions about when to prioritize organic versus paid investment, see this comparison of organic and paid ChatGPT options, and for the paid product-feed side specifically see this overview of ChatGPT ads for ecommerce.

What ChatGPT product recommendations are and where they appear

Per OpenAI’s help documentation, ChatGPT shows product options when a user’s query suggests shopping intent, such as looking for items to buy or comparing products. These results include imagery, product details, and purchase links, and for some eligible products and merchants they also include an Instant Checkout option, which is a way to complete the purchase inside ChatGPT without leaving the interface.

Shopping responses appear in ChatGPT on both mobile and web, so users can see recommendations across devices. Users can also save products they see into Favorites and organize them into folders in the ChatGPT Library, which keeps product picks accessible for later consideration.

A common misconception is that you can buy placement in these recommendations. OpenAI’s documentation states that product results "are selected independently by ChatGPT and are not ads, nor influenced by any OpenAI partnerships," so payment and partnerships do not control which products appear.

How ChatGPT chooses which products to show

When ChatGPT detects shopping intent, it selects products using several inputs. OpenAI states that selection considers the user’s query and context, including features like Memory and Custom instructions, so the model can align results with what it knows about the user’s preferences and constraints.

The system also uses structured metadata from first- and third-party providers, such as price and description fields from merchants and other commerce partners. In addition, it considers other third-party content and the model’s own responses generated before it looks at new search results, as well as OpenAI’s safety standards. Not all available products are shown, even if they are technically eligible.

Because OpenAI has not published any ranking weights beyond the named factors, there is no documented way to tune for a specific weight or to "game" the ranking. The practical implication for a merchant is that high-quality, accurate, and complete product metadata is the main controllable input the model uses when deciding which products fit a query.

How merchants are ranked and how prices are displayed

OpenAI states that merchant rankings are based on 4 factors: availability, price, quality, and whether the merchant is the maker or primary seller of the item. These signals are generated from merchant and product metadata, so the information merchants and providers supply is central to how they are ordered in the results. OpenAI also notes that ranking personalization will evolve, which means the system may adapt to user behavior over time.

Pricing in ChatGPT shopping responses comes from third-party providers rather than from OpenAI itself. The initial response typically shows the first listed merchant’s price, and OpenAI’s documentation clarifies that this is not necessarily the lowest price available for that product.

When there is a lower-price option, ChatGPT may label it "Best price" to highlight it for the user. Merchant pricing and shipping updates can be delayed in the interface, since they are sourced from external providers, so what a user sees may lag behind the merchant’s own site. This also corrects a common misconception that the price shown is always the lowest; OpenAI explicitly states that the first displayed price is usually just the first listed merchant’s price.

How merchants become eligible for ChatGPT shopping

As of October 2026, OpenAI documents 2 paths for merchants who want their products to be eligible for ChatGPT shopping responses. The first is for Shopify merchants, whose products are integrated automatically via Shopify Catalog. OpenAI’s help article states that this path requires no additional work from the merchant beyond being in Shopify Catalog.

The second path is for other merchants who are not covered by Shopify Catalog. These merchants can supply a direct product feed that follows the Product Feed documentation at developers.openai.com/commerce/specs, then apply at chatgpt.com/merchants. The product feed is a structured data file that contains product information such as titles, descriptions, prices, and other attributes in the format OpenAI specifies.

There is no published information in the research about acceptance criteria or timelines for the merchant application, so the only documented requirements are to provide a compliant product feed and to submit the application. For merchants evaluating effort, Shopify Catalog participation is the lower-friction path, while the direct feed path requires building and maintaining a feed to spec.

What merchants control versus what ChatGPT generates

Merchants control the structured metadata they supply, either through Shopify Catalog or through a direct product feed. OpenAI’s documentation notes that feed quality determines what the model has to work with, which means accurate, complete, and up-to-date fields like titles, descriptions, prices, and availability are the main levers a merchant can pull.

On top of that data, ChatGPT may generate simplified product titles and descriptions from third-party data. Feature labels such as "Budget-friendly" or "Most popular" are model-generated, not verified statements from merchants or OpenAI. This corrects the misconception that these labels are merchant-written or verified; OpenAI explicitly states they are generated by the model.

Review summaries are also model-generated from public reviews, and OpenAI notes that ratings are not verified by OpenAI. For clothing and accessories, there is a virtual try-on feature exposed as a Try on button, which uses the underlying product data but is part of the ChatGPT experience rather than something merchants directly configure.

Why you cannot pay your way into product recommendations

OpenAI’s help documentation states that product results "are selected independently by ChatGPT and are not ads, nor influenced by any OpenAI partnerships." This means there is no documented mechanism to buy placement or to have partnerships change how organic product recommendations are selected or ranked.

During the ads product-feed beta, OpenAI’s documentation on creating campaigns from product feeds specifies that feed products used for ads do not appear in organic conversations. Ads are a separate system from the organic product recommendations described here, and are covered in more detail in this overview of ChatGPT ads. This directly corrects the misconception that you can buy placement in ChatGPT’s product recommendations, which OpenAI’s documentation contradicts.

Misconceptions merchants should clear up before investing time

Several recurring misconceptions can distort how merchants evaluate ChatGPT shopping. The first is that you can buy placement in product recommendations, which OpenAI’s documentation contradicts by stating that results are independent, not ads, and not influenced by partnerships.

The second misconception is that the price shown in ChatGPT is always the lowest available. OpenAI clarifies that the initial response usually shows the first listed merchant’s price, and that the lowest-price option may instead be labeled "Best price," so users and merchants should not assume the first price is the cheapest.

A third misconception is that getting listed requires a complicated integration. For Shopify merchants, OpenAI states that integration through Shopify Catalog is automatic and requires no additional work, so the barrier is lower than many expect. For non-Shopify merchants, the documented requirement is to supply a product feed that follows OpenAI’s specs and to apply through the merchant application page linked above.

The final misconception is that labels and review summaries in ChatGPT shopping are merchant-written or verified. OpenAI states that labels like "Budget-friendly" and review summaries are model-generated from third-party data and are not verified statements, so merchants should focus on accurate source data rather than trying to author these elements directly.