Quick Answer: ChatGPT merchants get products into ChatGPT by applying at chatgpt.com/merchants to share a product feed, or automatically if they sell through Shopify or Etsy. Shopping is live for US users, purchases happen on your own site, and OpenAI charges no fee on them, as of 5 October 2026.
More people now start product research inside an AI chat than most retail teams realize, which is why your ChatGPT product feed now matters. Adobe measured AI-referred traffic to US retail sites up 393 percent year over year in the first quarter of 2026, and that traffic now converts better than most other sources.
If a customer queries ChatGPT about the leading choice in a certain category, the platform will reply with product cards that display pictures, cost, reviews, and purchase links. The query for any retailer, however, is straightforward: are the goods being searched for, part of their stock? This manual is the tactical variant where product data, feeds, and schema help any retailer get its stock to those results.
What is the ChatGPT merchants program?
The ChatGPT merchants program is OpenAI's application route for sharing a product feed with ChatGPT, run from chatgpt.com/merchants. The page, checked on 5 October 2026, sets out who needs to apply and what merchants keep control of.
| Question | What OpenAI's merchants page says (5 October 2026) |
|---|---|
| How to join | Apply with a form; if you have already applied, you are on the waitlist |
| Shopify or Etsy sellers | Catalog already integrated, no application or setup needed |
| Where shopping is live | For ChatGPT users in the US |
| Where the sale happens | Your site or app by default |
| Fees | None on purchases made on your site or app |
| Underlying standard | The Agentic Commerce Protocol, which today supports product discovery |
| Self-serve onboarding | A self-serve merchant platform is planned "later this year" |
Who should apply and who should not
Apply if you sell outside Shopify and Etsy and have a catalogue you can export as a structured feed. If your store runs on Shopify or Etsy, the application adds nothing: put the effort into the data quality and crawler access covered below.
The checkout question is settled for now. OpenAI scaled in-chat checkout back in March 2026, and the Instant Checkout status page explains what changed and which feed fields still matter.
What does ChatGPT Shopping look like in 2026?
ChatGPT Shopping is a collection of surfaces that help discover organic products in ChatGPT, rather than providing a paid ads service. The results are selected by OpenAI itself, and therefore you can't purchase your placement. This key fact should define your approach in its entirety.
The current status of the feature set is totally different, so don't pay attention to guides published before 2026. Here is the actual situation.
Where do products appear?
Three discovery experiences are live, primarily in the United States:
- In-answer product results. Inline cards with image, price, reviews, and a link out to the merchant, shown when a query has clear shopping intent.
- Shopping research. A slower, deep-research mode that asks clarifying questions and returns a buyer's guide after a few minutes.
- Visual product discovery. Browse products visually and compare options side by side, refining by budget or constraints in conversation.
All three direct the consumer to a website or mobile app for the actual purchase. The honest learning strategy for 2026 is to find out what ChatGPT has to offer followed by going to the retailer.
What changed, and what got scaled back?
OpenAI launched in-chat purchasing, called Instant Checkout, in late 2025 through the Agentic Commerce Protocol it built with Stripe. By March 2026 it had pulled back on that checkout push, with only around thirty Shopify merchants live, because keeping stock and price data accurate in real time proved hard.
The protocol was not terminated. It was further developed by OpenAI in the month of March, 2026 while complete product information of sellers like Target, Sephora, Best Buy, and Wayfair was included in ChatGPT. OpenAI's merchants page now says Shopify and Etsy catalogs are already integrated, so millions of stores reach ChatGPT without a separate feed.
Hence, the tale is not "obtain in ChatGPT." It is "be identified in ChatGPT with all the proper information about the product."
Are ChatGPT shopping results ads?
The outcomes of an organic product cannot be regarded as advertisements or involve any form of payment. OpenAI takes its decisions solely based on the relevance of the information.
A detail that needs to be pointed out here. OpenAI started testing its clearly-marked promoted product cards in the early 2026, clearly separated from the organic answer and not making an impact on it according to OpenAI.
This means that while organic recommendations do not involve any payments, ads represent a separate clearly marked type of content. Thus, make sure you do not inform your stakeholders that ChatGPT is ad-free as it is no longer correct.
How does OpenAI source products?
OpenAI pulls product details from three simultaneous sources: a crawler on your public product pages, first- and third-party structured data, and merchant feeds via the Agentic Commerce Protocol. Regarding the Shopify shops, the path of the feeds is automatic.
When there are multiple vendors for an item, OpenAI takes into consideration its rank in availability, price, quality, and also if one is the producer or seller of the item. Review statements are compiled from publicly published ratings. It is noted as no auction takes place, which provides the rating with valid structured information and third-party proof other than monetary reserves.
There is one detail that carries more weight than it may seem. Direct feeds outperform crawl-only listings, as scraped price and product data becomes obsolete and out of date, and outdated data is particularly what did not work in in-chat purchase. New, complete machine-readable product data serves as a sign of visibility and trust.
How to get your products into ChatGPT Shopping
The efforts are mainly centered on the performance of structured data hygiene. The effectiveness of OpenAI in crawling and authenticating you is ensured as well. Each strategy proceeds from the method of sourcing and ranking of products by OpenAI.
Add complete Product schema to every product page
Mark up each product with schema.org Product JSON-LD, including name, description, brand, an identifier like GTIN or MPN, and image. OpenAI's search bot reads this on-page metadata to extract product facts, and a GTIN lets it match your listing to third-party price and review data and dedupe you against other sellers.
Incomplete designs result in incomplete cards, and incomplete cards are inferior to complete ones. Our technical guide to schema markup for AI search addresses the topics that need to be accurate.
Keep Offer price and availability accurate and live
The Offer contains the accurate price, priceCurrency, and availability. Additionally, OpenAI has been ranking competing sellers according to their prices and stock availability. However, inaccurate stock information caused the project to fail badly when checking out.
See accuracy as a ranking element rather than a mere formality. When a listing shows that something is in stock when it is not, it undermines the trust the search engine is striving to preserve.
Expose real reviews with AggregateRating
Include AggregateRating and a personal Review with real rating figures and counts. Reviews are an unmistakable selection indicator and ChatGPT creates its likes and dislikes summaries from review data. No reviews mean that the engine would have no data to summarize.
Then go beyond your own domain. Earn corroborating reviews on independent sites and marketplaces, because OpenAI weighs first and third-party signals together, and independent corroboration is what breaks ties between sellers of the same item. That same corroboration principle sits at the heart of getting cited by ChatGPT for non-product queries too.
Submit a structured feed, or ride Shopify
The ranking of the direct feeds is better than that of crawl-only feeds, so it is worth it to send structured product feed through Agentic Commerce Protocol. If you sell through Shopify or Etsy, OpenAI says your catalog is already integrated, so most of this is handled for you.
Complete all the essential fields: stable item ID, clear title, straightforward description, brand, URL leading to 200, HTTPS image, price in universal currency code, availability value, GTIN, and reviews. Don't leave any fields blank or your ranking will be affected.
Reuse your Google Merchant feed
If you have sent a product feed to Google Shopping, then you are nearly done. The product surfaces at OpenAI use the same format as the retailers send to Google, thus a properly formatted feed from Merchant Center is the quickest way of getting the product data for ChatGPT.
At the same time, make your Merchant Center disapprovals right. The same flaws that can complicate Google will also ruin your ChatGPT listings.
Let OpenAI's crawler reach your pages
Check if OAI-SearchBot and ChatGPT-User are permitted in robots.txt. This typically consists of modifying robots.txt.liquid on Shopify so that allow rules are added above disallowed rules. When the search bot is blocked, your items would not be able to enter the discovery at all.
This is an extremely common silent hurdle, and just unblocking the site has taken stores from having no visibility to showing up in the results in no time. This follows the same crawler-access principle that decides how to rank in ChatGPT for ordinary answers.
Write attribute-rich titles and keep data consistent
Begin your product descriptions and titles with clear characteristics: dimensions, materials, application, resilience. Researching products for purchase relies on keywords like "best silent cordless stick vacuum for small flat," so the copy needs to be based on this. The same model of answer-first, extractable answers characteristics employed in AI search engines is a good approach to product descriptions, not just articles.
Last but not the least point to remember is that price and stock should match on your on-page schema, feed, and Google. If there are discrepancies in these systems, then you may face the stale errors called out by OpenAI; also, consistency can protect your ranking position in the search engine.
How do you keep the listings correct over time?
Treat it as a running check, not a launch task. Prices change, stock shifts, features get renamed, and rankings move, so a listing that was accurate at setup drifts on its own without anyone touching it.
Run the same buying prompts on a fixed cadence and confirm three things: that your products appear at all, that the price and availability shown match your feed, and that the review summary reflects real reviews rather than a thin sample. A wrong price inside an answer costs more than absence does, because it reaches the buyer with your name attached to it.
Feed accuracy is only half the job, though. Being the brand the engine reaches for in the first place is a category-level problem, covered in ChatGPT shopping for D2C brands, with the underlying method in answer engine optimization.
Automating the repeat is a separate decision again. Keeping the runs comparable is what decides whether the numbers mean anything, which measuring your AI share of voice covers.
The best AEO tools in 2026 compares what is available to do it with. Whatever you use, re-scan after every feed or schema fix ships, so you can see whether the listing actually changed.
