ChatGPT Shopping results shifted dramatically in July 2026: structured product feeds went from powering roughly 8% of recommendations to roughly 65% in a matter of days. If your products were visible on ChatGPT Shopping before July 10 and disappeared after, the reason is likely structural, not a content quality issue.
According to Profound, a firm that tracks its customers’ daily prompt runs on ChatGPT, the share of Shopping recommendations it classifies as “feed-integrated” (pulled from structured product data files rather than open web pages) jumped from 8.26% to 61.54% on July 10, 2026. This was observed across 1,757,723 tracked prompt runs in July. Profound’s analysis is observational, based on its own customer network, not ChatGPT’s total shopping traffic.
What Flipped on July 10?
Profound attributes the shift primarily to OpenAI’s release of GPT-5.6 on July 9, 2026, which the company said would roll out over the following 24 hours. However, as of Profound’s publication date, OpenAI has not confirmed this connection, and its ChatGPT release notes include no shopping-related updates for July 9 or 10.
Looking at a broader sample of 97,725 prompt runs from July 1 to August 24, 2026 (using a 10% sample counted once per prompt per month), Profound found feed retrieval (results sourced from structured data files) overtook web search retrieval in August. As of September 3, 2026, feed retrieval accounts for approximately 65% of the product recommendations Profound tracks.
Who Lost Visibility in ChatGPT Shopping, and Who Gained It?
The shift produced uneven results across Profound’s customer base. Of the 687 customers whose prompts consistently triggered Shopping results from July 7 to 12, 2026:
- 450 customers experienced a reduction in Shopping visibility of at least 33% when comparing July 7-9 against July 10-12.
- 67 customers saw a gain of that same magnitude.
- 517 customers in total experienced visibility changes of at least 33% in either direction.
A statistical model built on those 517 customers, measuring how much web-search retrieval each lost and how much feed retrieval each gained, explained 83% of the variation in visibility changes, according to Profound. The report frames this as an association within its own dataset, not a confirmed causal relationship.
The shift also concentrated traffic among fewer stores. The share of product references going to the top 10 merchants jumped from 22.5% to 41.8%. The number of unique merchants referenced dropped from 13,524 to 10,607, a decline of more than 20%.
Which Stores Are Already Connected to ChatGPT Shopping Feeds?
ChatGPT pulls structured product data through what OpenAI calls the Agentic Commerce Protocol (a standard that defines how stores share product information such as price, images, and descriptions directly with ChatGPT). OpenAI extended this protocol to product discovery on March 24, 2026.
As of the report’s publication date, two categories of sellers are already integrated:
- Shopify merchants: Product data is integrated automatically via Shopify Catalog, with no extra steps required. Profound estimates approximately 35% of all feed retrieval in July 2026 was Shopify-related, based on the close movement of their respective daily data series.
- Etsy sellers: Etsy catalogs are connected directly to ChatGPT Shopping.
Other retailers can apply for direct feed access, but OpenAI states that applicants are currently on a waitlist. Feeds can also come through providers like Salesforce and Stripe. OpenAI’s help center states that product results in ChatGPT Shopping “are selected independently by ChatGPT and are not ads, nor influenced by any OpenAI partnerships.”
ChatGPT Shopping is currently live for users in the United States, with OpenAI stating it will expand to more regions over time. Later in 2026, OpenAI plans to launch a self-serve platform that will let stores connect their own feeds without going through the waitlist process.
What This Means for AI-Search Visibility
The core issue is simple: if your products are not in a feed that ChatGPT can read, your odds of appearing in Shopping results have dropped sharply since July 10, 2026.
Think of it this way. Before July, ChatGPT found products the same way a search engine does: it crawled (scanned) your website and pulled product details from web pages. After July, it started relying mainly on structured product files, much like how Google Shopping works with Google Merchant Center feeds. A store without a connected feed is now less visible, regardless of how well-optimized its website is.
This is Hingewise’s read of the data:
- The concentration effect matters more than the average drop. The total number of unique merchants referenced fell from 13,524 to 10,607. It is not just that some stores lost visibility; fewer stores are referenced at all. The top 10 now capture 41.8% of references, up from 22.5%. That dynamic tends to compound over time as feed-connected stores accumulate more data signals and appear more frequently.
- Being on Shopify or Etsy is now a structural advantage, not just a convenience. These platforms gave their merchants automatic feed integration before most retailers understood this shift was coming. An independent retailer on a custom platform has to apply, wait, and then maintain a compliant feed to reach the same baseline.
- The next competition will not be about having a feed. It will be about feed quality. Profound anticipates the next phase of differentiation will focus on the specific fields a feed contains, not just whether one exists. This mirrors how Google Shopping evolved: early advantage went to sellers who submitted any feed; later advantage went to those with complete, accurate, and frequently updated data. ChatGPT Shopping looks to be on the same path.
- One data caveat is worth keeping in mind. Profound’s figures represent its own tracked prompt runs, not all ChatGPT Shopping traffic. The directional signal (feed retrieval rising sharply) is meaningful, but exact percentages should be read as estimates from one vendor’s network rather than platform-wide benchmarks.
Hingewise’s assessment: the July 10 shift looks less like a temporary algorithm fluctuation and more like a structural change in how ChatGPT sources product data, comparable to the long-term shift Google made toward Merchant Center feeds over organic product page results. Brands that treat feed integration as optional are likely to see their ChatGPT Shopping presence erode further as the self-serve platform opens and more feed-connected competitors enter.
What to Check Before Assuming Your Products Appear in ChatGPT Shopping
- Confirm whether your e-commerce platform already has a native feed connection to ChatGPT. Shopify and Etsy do by default; most others do not yet.
- If you are not on Shopify or Etsy, check OpenAI’s merchant page for current waitlist status for direct feed access.
- Review whether your existing feed providers (Salesforce, Stripe, or similar) are among the providers OpenAI currently supports.
- Audit your product data fields: price, description, availability, and images are the structured fields OpenAI’s help center specifically mentions as relevant inputs.
- If you have access to a prompt-tracking tool like Profound, check your Shopping visibility around July 10 to see how your store was affected by the shift.
- Watch OpenAI’s merchant page for announcements about the planned self-serve feed platform, which OpenAI says is coming later in 2026.
The model powering ChatGPT Shopping may also shift independently. OpenAI began rolling out GPT-6 Astra to a limited set of organizations on September 3, 2026, meaning the model handling Shopping prompts may not be uniform across all users right now. How feed retrieval behaves under GPT-6 Astra is still an open question worth monitoring in the weeks ahead.
Source: Search Engine Journal, “ChatGPT Shopping Results Lean Hard On Product Feeds,” Matt G. Southern. Underlying data from Profound, based on 1,757,723 tracked prompt runs in July 2026 and 97,725 prompt runs from July 1 to August 24, 2026.
This article was drafted with AI assistance from the sources cited above and checked against them before publishing. The "What this means for AI-search visibility" section is Hingewise's editorial analysis, built on the frameworks and positions we use in our GEO work. How we work: Editorial policy.