How AI shopping agents choose which products to show

There’s a lot of confident advice about ranking in AI shopping results. Very little of it cites anything. The companies running these channels have said less than the advice implies, but they have said some things, and they’re more useful than the guesses.

This post collects what they’ve published, as of September 2026.

Two steps: being available, then being picked

Getting shown happens in two stages, and they’re controlled by different companies.

1. Your product has to be available to the channel. For most AI channels, a Shopify store’s products arrive through Shopify Catalog. That part has published rules, which we list in Shopify Catalog requirements.
2. The channel decides what to show. Shopify is explicit that this part isn’t up to Shopify: the catalog “supplies product data that an AI channel can use, but each channel controls the final ranking, wording, and shopping experience” (Shopify Catalog).

Everything below is about the second step.

What OpenAI says ChatGPT looks at

OpenAI’s help article on shopping with ChatGPT search is the most specific public source from any AI channel. The points that matter for a merchant:

• Product results aren’t ads. “Product results are selected independently by ChatGPT and are not ads, nor influenced by any OpenAI partnerships.”
• It reads structured data from providers. Results draw on “structured metadata from first-party and third-party providers (e.g., price, product description) and other third-party content.” For Shopify stores, OpenAI says product data is “already integrated into ChatGPT through Shopify Catalog.”
• It can rewrite your product. “ChatGPT may generate simplified product titles and descriptions based on information it receives from third-party providers.”
• Reviews come from the open web. Review summaries “are based on reviews from public websites,” and “reviews and ratings are not verified by OpenAI.”
• Merchants are ranked on a few factors. When several merchants sell the same item, they’re “ranked based on factors like availability, price, quality, and whether they are the maker or primary seller of that item.”

The shopper’s own question and context also shape the result. OpenAI mentions “your query and context (such as Memory or Custom instructions).”

What Shopify’s listing-quality score rewards

Shopify’s admin has a search preview under Sales channels > Agentic that shows how your products might rank for a query. Next to each product is a Listing quality indicator, and Shopify documents what goes into it (Managing agentic storefronts):

• Description completeness: “Measures the word count of your product description.” Shopify’s reasoning is that AI channels “rely on descriptions to match products to natural-language customer queries.”
• Image coverage: the number of product images.
• Product reviews: average rating and review count, “calculated only from reviews that are verified by trusted sources.”
• Variant & option completeness: variants, options and in-stock availability. It “also checks whether your variant or option names contain acronyms or numbers that are hard for agents to understand.”
• Shop policy completeness: whether shipping, returns, refund and other policies are present.

Shopify adds a caveat worth keeping in mind: “Search relevance also depends on other signals such as popularity, customer engagement, and brand recognition.” It also warns that channels “often re-rank results based on their own logic,” so the preview is “a directional signal, not an exact prediction.”

What Google says about AI Mode

Google’s AI Mode and Gemini get Shopify products through Merchant Center, not directly from the catalog (Selling on Google AI Mode and Gemini). For the search side, Google’s guidance is short: “There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary” (AI features and your website).

Its listed best practices include “making sure your structured data matches the visible text on the page” and “checking that your Merchant Center and Business Profile information is up-to-date.”

What Microsoft says about Copilot

Microsoft’s January 2026 Copilot Checkout announcement notes that “product feeds from MMC help inform organic Copilot results,” where MMC is Microsoft Merchant Center (Microsoft Advertising blog). For Shopify stores, Shopify says you “don’t need to do anything” for Copilot to discover your products through the catalog (Selling on Microsoft Copilot).

What this adds up to

Nobody has published weights, and anyone claiming to know them is guessing. What the sources agree on is plain:

• Complete product data helps. Longer, clearer descriptions, more images, full variant data and filled-in policies are the things Shopify measures and the things channels read.
• Price and availability matter wherever the same product is sold by more than one store.
• Reviews count, but only ones a channel can find and trust.
• Consistency matters. Google asks for structured data that matches the page. OpenAI rewrites titles and descriptions from what it receives. The more your data agrees with itself, the less room there is for a rewrite to drift.

The part ranking advice skips

Every source above is about whether you get picked. None of them is about what the agent says once it picks you. OpenAI says it may generate its own titles and descriptions, and Shopify marks the catalog’s description, top features and selling points as fields its AI generates or enriches (What Shopify’s AI adds to your listings).

In our study of 192 top-ranked products, 89% had at least one claim in the catalog’s version the brand never made (how we checked, full study). Ranking well gets your product in front of the shopper. What the shopper hears about it is a separate question.

Hearsay answers that second question for Shopify stores: it reads what the catalog tells AI agents about each product and flags anything that contradicts your copy or that you never said. The free scan checks 5 products in under a minute.