Skip to content

AI Engines

Getting your Shopify products into ChatGPT shopping

Published

How ChatGPT surfaces Shopify products in 2026

ChatGPT surfaces Shopify products through Shopify Catalog. When a buyer asks a product question in ChatGPT, the assistant references Catalog product data and may show product cards. Tapping a card opens the merchant's storefront in an in-app browser or new tab — Shopify's docs describe ChatGPT as a 'discovery-focused referrer platform.' Direct in-chat checkout is not currently the default for ChatGPT.

The referrer framing is on Shopify's Agentic Storefronts help page2 and OpenAI's own product-discovery and Instant Checkout pages45. Catalog inclusion is the precondition; product-data quality determines what gets recommended; OAI-SearchBot determines whether ChatGPT can validate the product in real time.

The referrer model in detail

Verbatim from Shopify: ChatGPT is 'a discovery-focused referrer platform'2 — customers complete purchases on the merchant's store via in-app browser or new tab. The merchant's storefront then handles the checkout. ChatGPT-sourced sessions arrive on the PDP with chat.openai.com as the referrer.

The practical implication: ChatGPT-sourced conversion is storefront-conversion. A merchant who optimizes for in-chat checkout (Copilot, Gemini, Perplexity) doesn't automatically optimize for ChatGPT — those are storefront-PDP questions. Page speed, above-the-fold buy-box position, trust signals — all the classic storefront UX inputs matter for ChatGPT-sourced traffic.

What ranks inside ChatGPT shopping

ChatGPT recommendations draw on three inputs per OpenAI's own framing: training data (model's pre-existing knowledge), Bing search results (live web data), and the Catalog feed (when Shopify products are involved). For a Shopify merchant, Catalog data quality is the lever you control directly. Bing index quality matters because ChatGPT search references it. Training data matters in a slower-feedback loop.

Shopify's own AI optimization doc7 gives the field-level instruction: detailed specifications, comparison information, comprehensive descriptions, structured data, sizing/material/care, price/availability, alt text on images. Apply those across the top 20 products. The product data hub walks the field-by-field rewrite.

How to test ChatGPT surfacing of your store

Open ChatGPT (logged-in, US session). Ask three category-level queries that should surface your products. Note which products appear, which competitors appear, and what the in-app browser experience looks like when you tap a card. Repeat weekly for 4 weeks. This is the qualitative leg of the AI visibility audit; the quantitative leg is the Agentic Storefronts admin pane.

Examples by category: skincare → 'recommend a vitamin C serum under $50'; apparel → 'find a heavyweight cotton hoodie under $80'; supplements → 'recommend a creatine monohydrate with NSF certification'. Use queries a real buyer would type, not branded queries. Questions about the store itself, like return windows or shipping countries, show up somewhere else: the Knowledge Base query log lists what shoppers asked AI agents about your store and which agent received each question.

The ChatGPT surfacing checklist

Eight checks. Most stores pass the first four; the bottom four are where store-specific work lives.

  1. Catalog eligibility passes (see eligibility article).
  2. ChatGPT channel access is on in admin.shopify.com/agentic.
  3. robots.txt.liquid does NOT block OAI-SearchBot, ChatGPT-User, or GPTBot.6
  4. OAI-SearchBot is allowed (this is the bot that does live citation work).
  5. Top 20 products meet Shopify's seven AI-input categories.
  6. Storefront PDP loads under 2.5s LCP (ChatGPT-sourced sessions arrive on PDP).
  7. Storefront buy-box is above the fold with clear add-to-cart.
  8. Test queries weekly: confirm at least one product surfaces for one category query.

Keep reading