How to Appear in ChatGPT Shopping (2026)

You appear in ChatGPT Shopping by registering in OpenAI's merchant portal, passing business verification, and delivering a complete product feed to the secure endpoint OpenAI provides. Once approved, your catalog becomes eligible to surface when a user asks a shopping question that your products can answer.

Reporting from 2026 indicates OpenAI reviews and approves businesses for feed ingestion, then issues a dedicated secure SFTP endpoint with authentication credentials for feed delivery. Treat this like any other retail channel: the account exists to be maintained, not set up once and forgotten.

Because approval and portal details can change during this early phase, confirm the current onboarding steps in OpenAI's official merchant documentation before you build against any specific field or endpoint.

The single biggest lever is a complete, accurate, machine-readable feed. Reporting suggests ChatGPT accepts the same structured product file retailers already send to Google Shopping, in CSV, TSV, XML, or JSON, with updates as often as every 15 minutes.

Freshness matters as much as completeness. A feed that shows out-of-stock items or stale prices erodes trust with both the model and the shopper, so sync inventory and pricing on the tightest cadence your systems allow.

In-chat purchasing stalled and discovery took over as the primary shopping job inside ChatGPT. In late March 2026, OpenAI revamped its shopping experience to focus on helping users find and compare products, moving away from its earlier Instant Checkout push.

The economics drove the change. Walmart reportedly measured checkout inside ChatGPT converting roughly 3x worse than a click-through to walmart.com, even though ChatGPT drove about 2x the new-customer rate of search. The lesson: win the discovery moment, then convert on your own site.

Recommendations are largely decided by the depth and accuracy of your product data, so richer feeds win head-to-head comparisons. Because results are reportedly organic and not influenced by paid placement, the merchant with the more complete record tends to be surfaced.

Reviews appear to carry real weight in AI product selection, so expose ratings and review counts where the feed allows. Detailed attributes let the model match your product to a shopper's specific, conversational constraints instead of a generic keyword.

Because ChatGPT typically hands the buyer to your site, your product pages must convert the referral it sends. Match the promise made in the chat: the price, availability, and specifications a shopper saw should be identical on landing, or you lose the sale and the trust.

Reinforce the same machine-readable signals on-page with Product and Offer schema in JSON-LD, so the destination page corroborates the feed. A fast, clear product page also helps you appear in ChatGPT answers that cite live web content rather than only the feed.

Track referrals from ChatGPT as a distinct channel so you can prove the value of feed investment. Segment analytics by referrer and, where available, by UTM parameters on outbound links to separate assistant-driven visits from classic search.

Watch two numbers over a 30-to-90-day window: the share of new customers arriving from AI assistants and the conversion rate of that traffic once it reaches your site. Rising new-customer share with stable conversion is the signal your discovery optimization is working.