Your Shopify products are in ChatGPT by default: the checklist that actually gates them

If you run a Shopify store that sells to the United States, your products are almost certainly already discoverable in ChatGPT, and you did nothing to make that happen. Shopify’s own Help Center is blunt about it: “You don’t need to do anything for your products to be discovered by ChatGPT through Shopify Catalog.” The feature is “active by default for eligible stores.” So when a guide tells you to add an agents.md file, edit robots.txt, or buy an audit to “get into ChatGPT,” it is selling you a lever that is already pulled. If your products genuinely aren’t showing, the cause is almost never a missing optimization. It is a short eligibility checklist, and most of it is boring store hygiene you can fix in an afternoon for zero dollars.

I write for a lab that builds AI-visibility tooling, so I read these vendor pages the way a plumber reads a leak report. The pattern in the “why aren’t my products in ChatGPT” genre is consistent: a lot of energy spent on things that don’t gate discovery, and near silence on the handful of settings that actually do. Here is what the primary docs say, quoted, and what to check in order.

“On by default” is the whole point

Shopify products reach ChatGPT through Shopify Catalog, the same catalog that also feeds Microsoft Copilot and Google’s AI surfaces. For an eligible store this is opt-out, not opt-in. From the Shopify Help Center: the feature is “active by default for eligible stores,” and you can remove access from your admin if you want to leave. There is even a lag on the way out: removing Catalog access “can take up to 7 days before your product data is no longer being shared.” Nothing about that pipeline is something you switch on with a file in your theme.

That single fact reframes the whole question. The job is not “how do I get in.” You are in. The job is “which eligibility gate is quietly holding me out,” because the pipe only carries stores and products that clear a specific bar.

The store-level gates

Three settings decide whether your store qualifies at all. Miss one and no amount of per-product work matters.

1. You sell to the United States. The Help Center: “Your store must sell to customers in the United States, however, your store can be based outside the United States.” So a store run from Delhi or Berlin is fine, as long as the US is one of your markets. If you have never enabled a US market or US shipping, that alone keeps you out.

2. Your three core policies are published. This is the one most stores fail without realizing it. Shopify requires, verbatim, that “You must have your Terms of service, Privacy policy, and Return and refund policy completed in Settings, Policies.” A store that launched fast and never filled in its refund policy is not eligible, regardless of how good the products are. Two minutes in Settings, Policies clears it.

3. You have accepted the terms. Per the agentic storefronts overview: “To use Shopify’s agentic storefronts, you must read and agree to the Shopify Agentic Storefronts Supplemental Terms of Service.” If someone on your team opted the store out, or never accepted, that is a deliberate switch worth checking before you blame the algorithm.

The product-level gates

Once the store qualifies, individual products can still be filtered out. Four conditions do most of the filtering.

Unlisted products are hidden. A product whose status is set to Unlisted is excluded from agentic discovery. It is easy to leave a product Unlisted after a sale or a seasonal pull and forget it.

B2B-only and gated products are excluded. Shopify’s product-discovery docs exclude “Products that you publish only to B2B markets, or in product catalogs assigned to specific companies,” along with anything behind a password page or a login wall. If your catalog is wholesale-first, most of it will not surface.

Prohibited categories never show. The line is explicit: “If you have any products that are included in OpenAI’s Prohibited products and services, then they don’t display in ChatGPT product discovery results.” No setting on your end changes that.

The basic attributes have to be there. Catalog syndicates products with “title, description, options, images, price, availability, and other key attributes.” A product missing a price, an image, or availability is not a candidate. This is the one place the “clean up your data” advice is actually correct, but at the level of presence, not polish.

The robots.txt distraction, from Shopify’s own mouth

Because so much advice fixates on unblocking crawlers, it is worth quoting the one line that settles it. Shopify: “Blocking AI crawlers at the robots.txt or network layer affects only open-web discoverability. It doesn’t stop your product data from being sent by Shopify Catalog.” Your product data travels as a structured feed, not as a page a bot scrapes. Editing robots.txt moves your crawlable content, not your catalog. That is the same split I wrote about with OpenAI’s crawlers: the pipe that carries products is a feed, and the crawler you were told to unblock is a different mechanism entirely.

Eligibility is not recommendation

One honest caveat, so this reads as the full picture. Clearing the checklist makes your products eligible to appear. It does not make an agent choose them over a competitor when a shopper asks an unbranded question. That second problem, whether your titles, descriptions, and attributes give an agent a reason to name you, is the harder one, and it is genuinely yours to solve. But it is a separate problem, and it is the wrong thing to work on while a missing refund policy is quietly keeping your entire store out of the pipe. Fix eligibility first. It is faster, it is free, and it is the actual blocker for most stores that “can’t get into ChatGPT.”

If you want to see which of these gates your store trips before you go setting by setting, our free scan at pollen.elytron.in checks the store hygiene and product-data completeness these surfaces read.