How Is ChatGPT Turning Into a Shopping Platform With Product Feeds and Checkout?

How did ChatGPT's shopping features develop?

The main steps, from OpenAI's announcements and merchant pages:

OpenAI's merchant page now describes the flow simply. Shoppers browse rich results with images, prices and key details in ChatGPT, then complete the purchase on the merchant's site or app. Apps inside ChatGPT remain an option for larger merchants that want more control over the shopping experience.

How do product feeds work in ChatGPT?

A product feed is a structured file listing every product you sell, with its price, availability, images and details. OpenAI's feed specification sets out one row per purchasable item or variant, with nine required fields: item ID, title, description, product URL, brand, seller name, image URL, price and availability. Titles are limited to 150 characters and descriptions to 5,000, in plain text. Variants are grouped with a group ID, and availability must be stated as in stock, out of stock, pre-order, backorder or unknown.

Feeds are optional in one sense. OpenAI's FAQ says you do not need one if ChatGPT already crawls your site, but a feed gives you more control over how products appear and keeps information accurate and current. It supports common formats and connections, including SFTP, APIs and leading commerce platforms and feed providers. The standard format currently targets the US. Structuring the underlying data is covered in how ecommerce brands can structure product data for AI recommendations.

Does ChatGPT Shopping use Google Merchant Center feeds or a separate product source?

A separate source, but one that can accept a Google-style file. ChatGPT does not read your Google Merchant Center account. Its product data comes from its own feed programme, from catalogues connected through Shopify and Etsy, from crawling the web and, according to a 2026 study, largely from Google Shopping results that ChatGPT queries in the background, which are built from Merchant Center feeds. OpenAI's specification does, however, include a Google-compatible format. It accepts a tab-delimited or comma-delimited file using Google's column names, such as id, link and image_link, once OpenAI confirms that format for your registered feed. XML and g:-prefixed names are not supported on that path, and the registered merchant name is used as the seller on every row. In practice, a retailer with a clean Merchant Center feed can often reuse much of it, exported in the supported format, rather than rebuilding the catalogue.

Can optimizing Google Shopping feeds improve ChatGPT visibility?

Yes, and more directly than the shared data work suggests. Accurate titles with the product name and variant, factual descriptions, valid GTINs, correct availability and prices, and good images help any system identify and match products. OpenAI's specification warns against inventing identifiers and treats omitted shipping prices as unknown rather than free, so gaps in a feed carry into ChatGPT too.

The final ranking is OpenAI's own, though. OpenAI said at launch that product results are organic and unsponsored, ranked on relevance to the user. When several merchants sell the same product, it named availability, price, quality and whether the merchant is the primary seller or maker as factors in which merchant appears first. A strong Google Shopping feed gives you clean data to submit. Google's ranking matters more than you might expect: a 2026 Search Engine Land analysis with Peec AI found that 83% of ChatGPT carousel products matched a product in Google Shopping's top 40 results, so a well-ranked Google Shopping listing is one of the main routes into ChatGPT's carousel, alongside reviews and editorial coverage. Why rival stores can still win is covered in why ChatGPT recommends competitors instead of your online store.

Do ChatGPT shopping features increase visibility for AI-recommended stores?

They increase the chance of appearing accurately, not the ranking itself. OpenAI's merchant page says sharing feeds improves how products appear, with complete and current data, and makes more of a catalogue available to appear in results. That matters most for large catalogues, fast-changing prices and stock, and products the crawler describes poorly.

What a feed does not do is buy placement in the organic results. OpenAI says answers are not influenced by ads, and paid placements are labelled and shown separately. The clearest benefit is for stores that are already being recommended: a feed keeps their prices and stock right when shoppers are close to buying. For a store that is not recommended at all, the usual causes lie elsewhere: weak third-party coverage, thin product pages, or a price that looks worse than rivals'.

How will ChatGPT Shopping affect Google Shopping SEO?

It adds a channel rather than replacing one, at least for now. Product searches in ChatGPT are still a small share of its use: an NBER study of ChatGPT conversations found that about 2% of messages were about purchasable products. Google also handles far more searches overall. Google is moving in the same direction too, with its Universal Commerce Protocol, announced in January 2026, enabling checkout on eligible listings in AI Mode and Gemini for US retailers.

The practical effect on Google Shopping SEO is that product data work now serves several destinations. The same catalogue feeds Google Merchant Center, ChatGPT, and agentic storefronts from platforms such as Shopify. Teams that treat the feed as a shared, well-maintained source of truth will serve all of them. Teams that tune titles only for Google's ranking will find less of that work carries over.

What should retailers do now?

1. Apply to share your product feed through OpenAI's merchant page, unless you sell through Shopify or Etsy, where catalogues are already connected.
2. Clean the feed you already have: titles with variants, factual descriptions, valid GTINs, availability and shipping stated, not left blank.
3. Check whether your Merchant Center file can be exported in the Google-compatible format OpenAI accepts.
4. Keep product pages detailed, since ChatGPT still crawls them and draws on reviews and editorial coverage.
5. Track product appearances in ChatGPT's shopping results, and compare prices and attributes shown with your catalogue. Tools for this are compared in Profound vs Evertune for retail product recommendation monitoring.
6. Plan for merchant-owned checkout: make sure landing pages from ChatGPT load the right variant and price.

What might this look like in practice?

Consider an illustrative outdoor clothing retailer, Ilvoway, with 4,000 variants and a mature Merchant Center feed. It applied to share its feed with OpenAI and, once accepted, submitted an export in the Google-compatible format, after fixing 300 rows with missing GTINs and blank shipping prices. It then checked 40 shopping prompts in ChatGPT each week. Jackets that had appeared with outdated prices began showing current prices and sizes, and clicks from ChatGPT landed on the correct variant pages, where the purchase was completed.

What do people ask about ChatGPT shopping?

Can shoppers still buy inside ChatGPT?

OpenAI says it is moving away from standalone Instant Checkout; purchases are completed on merchant sites or apps.

Does OpenAI charge merchants for sales from ChatGPT?

Its merchant page says there are no fees on purchases that start in ChatGPT and are made on your site or app.

Do I need a product feed?

Not strictly, since ChatGPT crawls sites, but a feed gives more control and more accurate, current product data.

Can I use my Google Merchant Center feed?

Not directly, but OpenAI accepts a Google-compatible file format once it confirms it for your registered feed.