How to Get Your Products into ChatGPT Shopping Recommendations

ChatGPT shopping has quietly turned a chat window into a storefront. Ask ChatGPT for a product recommendation and it no longer hands you ten blue links to sort through. It names a short list of specific items, shows price and availability, and, for a growing set of merchants, lets you buy without leaving the conversation. For brands, that shift is bigger than it sounds. When the answer is three products instead of a page of results, being one of the three is the whole game. This guide covers what ChatGPT shopping is, how it decides what to recommend, whether the traffic is worth chasing, and the concrete steps to get your catalog in front of it.

What is ChatGPT shopping, and how does it work?

ChatGPT shopping is the set of product-discovery and checkout features built into ChatGPT. When you ask a buying question, ChatGPT returns product recommendations pulled from merchant catalogs, complete with images, prices, and links. OpenAI describes these results as organic and unsponsored: ranked purely on relevance to the request rather than on who paid to be there.

The buying half arrived on September 29, 2025, when OpenAI launched Instant Checkout: the ability to buy directly inside the chat. It opened with US Etsy sellers, with more than a million Shopify merchants named as next in line, and early brands like Glossier, SKIMS, Spanx, and Vuori. At launch it handled single-item purchases, with multi-item carts on the roadmap: a deliberately narrow start. On March 24, 2026, OpenAI expanded the discovery side, bringing large retailers (Target, Sephora, Nordstrom, Lowe’s, Best Buy, The Home Depot, and Wayfair among them) into the system so their catalogs surface in ChatGPT.

It is worth remembering this is OpenAI’s second run at agentic shopping; earlier attempts to send shoppers off to retailers were clumsy, and the current feed-and-checkout model is the correction. That matters for planning: the surface is still early and shifting, which is exactly why the brands that show up now tend to define the category before it gets crowded. ChatGPT shopping sits inside the wider shift toward AI search and LLM visibility, and if you want the broader picture of how models pick what to cite, our guide on how LLMs choose their sources covers the mechanics.

How does ChatGPT decide which products to recommend?

OpenAI has been clear about the inputs even while keeping the exact algorithm private. Recommendations are ranked on relevance, and the signals that feed relevance include product availability, price, quality, and merchant standing. In plain terms: an in-stock, fairly priced, well-reviewed product from a credible seller has a real shot; a thin listing with stale data does not.

The raw material is your product feed. Merchants share structured feeds and promotions so their full catalog is represented, and OpenAI supports several delivery paths, including third-party providers like Salesforce and Stripe. Shopify stores get a shortcut. Their product data is already wired into ChatGPT through Shopify Catalog, so the integration is largely automatic. The takeaway is that ChatGPT is reading structured commerce data, not scraping your marketing copy, which means the quality and freshness of your feed is doing most of the work.

Freshness deserves its own line. A feed that says “in stock” when the item sold out yesterday, or lists last month’s price, is worse than no feed. It puts a bad answer in front of a buyer and erodes the merchant-quality signal OpenAI weighs. The mechanics here rhyme with everything else in AI visibility: models reward clear, specific, verifiable data and quietly demote vagueness. A product page with real specifications, honest reviews, and complete attributes is legible to a language model in a way that a glossy hero image and three adjectives are not.

Does traffic from ChatGPT actually convert?

This is where the skeptics get quiet. A Visibility Labs study reported by Search Engine Land tracked 94 seven- and eight-figure ecommerce brands across all of 2025, comparing 9.46 million non-branded organic sessions against roughly 135,000 ChatGPT referral sessions. ChatGPT traffic converted at 1.81% versus 1.39% for non-branded organic, a 31% higher conversion rate, and earned $3.65 in revenue per session against $3.30. People arriving from a ChatGPT recommendation show up further along in the decision, because the model already did the shortlisting.

There is a nuance worth keeping honest: average order value actually ran lower on ChatGPT, $204 versus $238 for organic, so the win comes from conversion rate and revenue per session, not bigger baskets. And the absolute numbers are still small. Across the full year, ChatGPT drove about $474,000 in attributed revenue against $32.1 million from non-branded organic for the same brands.

The signal is the slope, not the size.

The volume story is just as telling. Across those same brands, ChatGPT sessions grew 1,079% year over year (from 1,544 sessions in January to 18,202 in December) while non-branded organic grew 17%. The gap between the two channels narrowed from 70:1 to 47:1 in a single year. Small base, steep curve. One catch: attribution is genuinely hard here, because a shopper can research inside ChatGPT and convert later through a branded search or a direct visit, so referral numbers likely undercount ChatGPT’s real influence. For a longer look at what that decline in classic search traffic means, see why your rankings are fine but your traffic isn’t.

How do you get your products into ChatGPT shopping?

There is no button that says “rank me.” What there is: a feed and a standard. The practical path breaks down into a few moves.

Get a clean product feed in front of it. If you are on Shopify, confirm your catalog is enabled for ChatGPT through Shopify Catalog. If you are not, submit a structured feed (directly or through a supported provider such as Salesforce or Stripe) and treat it as a living document. Fix the four relevance signals. Keep availability accurate, price competitive and current, product quality reflected in real reviews, and your merchant reputation clean; those are the levers OpenAI has named. Write listings for a machine that reads specifics. Titles, attributes, materials, sizes, and use-cases in structured fields beat vague marketing prose. Apply where an application is required. Instant Checkout onboarding runs through OpenAI’s merchant program, and Stripe customers can enable agentic payments in about one line of code. None of this is exotic; it is feed hygiene plus a decision to participate.

The brands that treat their catalog as first-class data will keep winning the shortlist.

What about Instant Checkout and agentic commerce?

Instant Checkout is the piece that removes the last click. A shopper can complete a purchase inside ChatGPT, and the plumbing underneath is the Agentic Commerce Protocol (ACP): an open standard OpenAI co-developed with Stripe and open-sourced, so any AI agent, shopper, and business can transact on shared rails. PayPal joined the effort in October 2025 to help power Instant Checkout and agentic commerce as well, which tells you the payments industry is treating this as infrastructure, not a novelty.

The economics are worth noting. Checkout is free for shoppers, and merchants pay a small fee only on completed purchases: closer to a marketplace commission than an ad spend. That is a different bargain than paid search, where you pay for the click and hope it converts. Here you pay when it works.

For merchants weighing agentic commerce against their paid budget, this reframes the whole calculation.

It also changes what a “conversion” looks like operationally. When the purchase happens inside ChatGPT, your checkout, your upsell flow, and your post-purchase email sequence are no longer the first touch, the protocol is. That is a trade: you give up some control over the buying moment in exchange for reaching a shopper at the exact instant of intent, on a surface where a competitor might otherwise get named instead. For most catalogs the math favors participating, but it is worth deciding deliberately rather than by default.

How is this different from Google Shopping and SEO?

Three differences matter. First, placement is earned, not auctioned: OpenAI has said results are organic and unsponsored, so for now you cannot simply outbid your way in the way you can on Google Shopping. Second, the surface is narrow. Google gives a shopper a grid of options and a page of ads; ChatGPT often gives a handful of named products. Fewer slots means the cost of not being cited is higher. Third, the ranking currency is relevance and data quality rather than backlinks and ad rank, which rewards clean feeds and honest listings over aggressive link building.

That does not make traditional SEO irrelevant: the same authority and clarity that earn citations in AI answers also help here. It does mean ecommerce teams need a second discipline sitting next to search: feed quality, structured product data, and participation in agentic commerce standards. The brands treating ChatGPT shopping as a channel today are building a head start while the surface is still uncrowded and unsponsored.

One more practical difference: measurement. Google Shopping hands you tidy campaign dashboards; ChatGPT gives you referral traffic that is easy to under-credit and a checkout that lives inside someone else’s app. Teams that win here set up the tracking early: tagging ChatGPT referrals, watching branded-search lift after AI-heavy weeks, and treating the small direct number as a floor rather than a ceiling. If you cannot see the channel, you will underinvest in it right as it is compounding.

Sources

• OpenAI — “Buy it in ChatGPT: Instant Checkout and the Agentic Commerce Protocol,” September 29, 2025.
• OpenAI — “Powering Product Discovery in ChatGPT,” March 24, 2026.
• Search Engine Land — “ChatGPT ecommerce traffic converts 31% higher than non-branded organic search” (Visibility Labs data), 2026.
• Semrush — US keyword volume and CPC data for “chatgpt shopping,” “chatgpt instant checkout,” retrieved July 2026.
• Stripe Newsroom — “Stripe powers Instant Checkout in ChatGPT and releases the Agentic Commerce Protocol co-developed with OpenAI,” 2025.
• PayPal Newsroom — “OpenAI and PayPal Team Up to Power Instant Checkout and Agentic Commerce in ChatGPT,” October 28, 2025.