ChatGPT Shopping: how to get your products recommended

ChatGPT shopping now shows visual product cards, prices, and store links directly inside the chat, and it does so for a user base that OpenAI and reporting put in the hundreds of millions of weekly users, reported at around 800 million weekly active users in late 2025. For merchants that is a new discovery surface with its own rules. The good news is that most of those rules are things you already control: your product feed, your structured data, and what third parties say about you. This guide breaks down how the AI engines choose products, which signals move the needle, and the one dependency that most optimization advice skips.

How does ChatGPT decide which products to recommend?

ChatGPT builds product recommendations by running your shopping question as a set of searches across the web, pulling structured product data, and ranking items on how well they match your intent. It is not a static catalog. When you ask for the quietest cordless stick vacuum for a small apartment, the model reformulates that into product-shaped queries, gathers candidates, and assembles a comparison.

The signals that decide inclusion are consistent across OpenAI's own documentation and independent testing: query relevance, structured product data on the page and in the feed, live price and availability, review quantity and rating, and what authoritative third-party sources say about the product. This is the same discovery logic covered in our generative engine optimization guide, applied to commerce. If a listing is missing structured data or contradicts its own page, the model treats it as low confidence and reaches for a cleaner competitor.

Are ChatGPT shopping results ads or organic?

They are organic. OpenAI states that product results are unsponsored and ranked purely on relevance to the user, with no paid placement, per its announcement of Instant Checkout. There is no bid, no ad slot, and no way to pay to jump the line the way you can in Google Ads.

That changes the playbook. Because you cannot buy visibility, the only levers are quality and clarity of information: accurate data, verifiable trust signals, and content the model can quote with confidence. It also means the discipline is closer to answer engine optimization than to paid search. You are optimizing to be the best-supported answer, not the highest bidder.

Why does your Google Shopping feed matter more than a ChatGPT feed?

This is the information-gain point most guides miss. They tell you to apply to OpenAI's merchant program and submit a dedicated ChatGPT feed, which is worth doing, but they skip where the products actually come from. Independent testing of thousands of ChatGPT product carousels found that 83% of carousel products matched Google Shopping's top-40 organic listings for the corresponding query, with roughly 60% coming from the top 10. A separate Semrush analysis reached the same conclusion: ChatGPT leans heavily on Google Shopping organic data to build its picks.

The takeaway is direct. Your Google Merchant Center organic feed is, for now, the primary input that decides whether ChatGPT can even see your product. Ranking well in Google Shopping organic results is a prerequisite, not a separate project. The same feed hygiene we cover in ecommerce product feed optimization pays off twice.

What product schema and feed fields does ChatGPT read?

ChatGPT parses your page's structured data before it trusts your listing. Schema.org Product, Offer, and AggregateRating markup is the baseline for inclusion. When the markup is missing, malformed, or contradicts the visible page, the model discounts the listing. A complete, valid block does the opposite: it hands the model clean facts it can render without guessing. Our product schema guide covers the exact JSON-LD to ship.

Richer listings win. Attribute depth matters, so materials, compatibility, sizing, and care details all give the model more ways to match a specific query. Here are the fields that carry the most weight for AI shopping surfaces:

• Identity: product title, brand, GTIN or MPN, and a canonical product URL.
• Commercial: current price, currency, and live availability status.
• Trust: review count and average rating, both product-level and store-level.
• Descriptive attributes: category, material, size, color, and compatibility.
• Media: a clean primary image URL that resolves and matches the item.

Keep the feed, the schema, and the visible page in agreement. Contradiction between them is one of the fastest ways to get filtered out.

How do reviews and third-party authority affect recommendations?

Reviews are not a nice-to-have here, they are a ranking input. A Product schema block with no AggregateRating or Review property tends to sit below a competitor that includes review markup, even when the underlying product is identical and cheaper. Review count and average rating function as trust signals the model reads directly.

Off-page signals matter just as much. ChatGPT weighs what independent sources, roundups, and review sites say about a product, because those corroborate the merchant's own claims. This is where commerce meets classic authority building: earning mentions on the pages the AI engines already read. If you want to be the recommended option, your reputation has to exist outside your own domain, which is exactly the logic behind review management for SEO and getting cited by AI.

What are Instant Checkout and the Agentic Commerce Protocol?

In late 2025 OpenAI launched Instant Checkout, which lets US users buy products without leaving the chat, starting with Etsy sellers and with over a million Shopify merchants including Glossier, SKIMS, Spanx, and Vuori set to follow. The market reaction was immediate: Etsy shares jumped about 16% on the announcement.

It runs on the Agentic Commerce Protocol, an open standard co-developed by Stripe and OpenAI. Merchants keep control of orders, payments, and fulfillment through their existing systems, pay a small fee only on completed purchases, and the fee does not affect prices or influence product ranking. Two practical notes for 2026: Shopify and Etsy merchants are eligible without a separate application, everyone else applies through OpenAI's merchant program, and the rollout has been iterative, so treat exact availability as a moving target rather than a fixed spec.

How do you write product content ChatGPT can quote?

The model favors clean, factual, use-case-rich descriptions over marketing filler. Write the way a knowledgeable salesperson answers a specific question: who the product is for, what problem it solves, what it is made of, and where it fits. Vague superlatives give the model nothing to match against a query like best gift for a coffee lover under 50 dollars.

Lead each product page with a direct, self-contained summary, then support it with concrete attributes and an honest FAQ that mirrors how buyers actually ask. That structure is the core of strong product page SEO, and it doubles as citable material for the AI engines. On Shopify specifically, feed and template defaults can either help or quietly sabotage this, which we unpack in the Shopify SEO guide.

ChatGPT Shopping vs Google Shopping: what is different?

The two surfaces overlap heavily on inputs but behave differently for shoppers and merchants. Here is the practical comparison:

The strategic implication: winning Google Shopping organic is the foundation, and clean schema plus reviews are what let ChatGPT confidently promote you from candidate to recommendation.

A practical ChatGPT shopping optimization checklist

If you want one prioritized list to act on this quarter, work top to bottom:

• Fix the Google Shopping foundation. Get products into Merchant Center with accurate titles, GTINs, price, and availability, and earn organic visibility there first.
• Ship complete Product, Offer, and AggregateRating schema that matches the visible page exactly.
• Populate every relevant attribute. Depth beats brevity for query matching.
• Build real reviews and expose the rating in both markup and content.
• Earn third-party mentions on roundups and review sites the AI engines read.
• Rewrite thin descriptions into factual, use-case-led copy with a self-contained summary up top.
• Apply for merchant eligibility if you are not already on Shopify or Etsy, and keep the feed fresh.

Disclosure: this guide is written by the team behind Sorank, which builds tooling in this space. The steps above work with any stack.

Frequently asked questions

Related guides