{"id":619767,"date":"2026-09-09T23:02:08","date_gmt":"2026-09-09T23:02:08","guid":{"rendered":"https:\/\/www.newsbeep.com\/ie\/619767\/"},"modified":"2026-09-09T23:02:08","modified_gmt":"2026-09-09T23:02:08","slug":"chatgpt-shopping-results-lean-hard-on-product-feeds","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/ie\/619767\/","title":{"rendered":"ChatGPT Shopping Results Lean Hard On Product Feeds"},"content":{"rendered":"

Profound<\/a> data shows the share of ChatGPT Shopping product recommendations it classifies as feed-integrated jumped from 8.26% to 61.54% on July 10, across 1,757,723 tracked prompt runs in July.<\/p>\n

The company runs its customers\u2019 prompts on ChatGPT every day, and from the network logs, it labels each product based on whether it comes from web search or what it calls feed-integrated retrieval. This analysis is observational, so all the figures below relate to those prompt runs rather than ChatGPT\u2019s entire shopping traffic.<\/p>\n

Store visibility moved around July 10. Out of the 687 customers in the sample, 450 experienced at least a one-third reduction in their Shopping visibility when comparing July 7 to 9 with July 10 to 12, while 67 saw an increase of that same magnitude.<\/p>\n

Feed Retrieval Overtook Web Search In Profound\u2019s Tracking<\/p>\n

Feed-sourced recommendations went from being a small part of the tracked runs before July 10 to becoming the majority afterward. This change is based on a daily series sampling 50% of July prompt runs, counting each prompt once per day.<\/p>\n

When looking at a broader view of 97,725 prompt runs from July 1 to August 24, with a 10% sample counted once per prompt per month, we see feed retrieval overtaking web search in August. As of September 3, the report<\/a> indicates that feed retrieval accounts for about 65% of the product recommendations it tracks.<\/p>\n

In addition, OpenAI released GPT-5.6<\/a> on July 9, stating that the rollout would happen over the next 24 hours. The report<\/a> attributes most of the July changes to this release. However, as of the publication date, OpenAI hasn\u2019t made this connection, and its ChatGPT release notes<\/a> do not include any shopping-related updates on July 9 or 10.<\/p>\n

Visibility Swung Across The July 10 Break<\/p>\n

In July, all 687 customers in the visibility sample actively tracked prompts that triggered Shopping each day and consistently had at least one product card pointing to their own products from July 7 to 12. During these same days, 517 customers experienced visibility changes of at least 33% in either direction.<\/p>\n

A statistical model based on these 517 customers, which analyzed how much web-search retrieval each lost and how much feed retrieval each gained, was able to explain 83% of the variation in their visibility changes. This finding is presented by the report as an association within its own dataset.<\/p>\n

Additionally, retrieval activity focused on a smaller group of stores during this period. The share of references to the top 10 increased from 22.5% to 41.8%, while the number of unique merchants referenced dropped from 13,524 to 10,607, a decrease of over 20%. The report suggests that feed retrieval drew from a narrower pool of compared to web search.<\/p>\n

The report also estimates that approximately 35% of all feed retrieval in July was related to Shopify, based on the close movement of the respective daily series.<\/p>\n

What OpenAI\u2019s Pages Say<\/p>\n

OpenAI\u2019s Help Center<\/a> says ChatGPT picks products using structured product data, such as price and description, from data providers and from stores themselves. The page also says: \u201cProduct results are selected independently by ChatGPT and are not ads, nor influenced by any OpenAI partnerships.\u201d<\/p>\n

Shopify stores already have their product data integrated via Shopify Catalog, with no extra steps needed, as indicated on the page. Other retailers can request direct feed access, but OpenAI\u2019s merchant page<\/a> states applicants are on a waitlist. It also mentions Etsy catalogs are connected, and Shopping is currently live for ChatGPT users in the U.S. These feeds operate through the Agentic Commerce Protocol, OpenAI\u2019s standard for sharing product data with ChatGPT.<\/p>\n

The protocol was extended to product discovery<\/a> on March 24, and feeds can come from providers like Salesforce and Stripe. The pages do not specify how often ChatGPT relies on feed data versus the open web, and OpenAI has not published that information as of publication.<\/p>\n

Why This Matters<\/p>\n

Profound\u2019s data suggests that how your product data reaches ChatGPT may matter more to Shopping visibility than it did earlier this year. Shopify and Etsy sellers already have their catalogs connected to ChatGPT. Everyone else has the direct-feed application, which OpenAI<\/a> says is waitlisted for existing applicants, or one of the feed providers OpenAI supports<\/a>.<\/p>\n

The data doesn\u2019t show you how your products rank once ChatGPT is pulling from feeds, and it only covers one vendor\u2019s tracked prompts instead of actual shopper sessions.<\/p>\n

Looking Ahead<\/p>\n

OpenAI<\/a> says Shopping will reach more regions over time. Later this year, a new self-serve platform is planned to let stores connect feeds on their own. Profound<\/a> anticipates that the next competition will focus on the types of fields a feed contains, rather than just having one. Additionally, OpenAI started rolling out GPT-6 Astra to a limited set of organizations on September 3, so the specific model used to answer prompts might vary when comparing July\u2019s results.<\/p>\n","protected":false},"excerpt":{"rendered":"Profound data shows the share of ChatGPT Shopping product recommendations it classifies as feed-integrated jumped from 8.26% to…\n","protected":false},"author":2,"featured_media":619768,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[61,60,80],"class_list":["post-619767","post","type-post","status-publish","format-standard","has-post-thumbnail","category-technology","tag-ie","tag-ireland","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts\/619767","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/comments?post=619767"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts\/619767\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/media\/619768"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/media?parent=619767"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/categories?post=619767"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/tags?post=619767"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}