ChatGPT Product Discovery Just Made Your Merchant Feed a Shortlist Asset
OpenAI's product discovery update means merchant feeds now influence whether products appear in ChatGPT shopping results. Here's the operator checklist.
OpenAI's ChatGPT product discovery update changes the work from "rank my category page" to "make my catalog machine-readable enough to be shortlisted." OpenAI says merchants can share product feeds through the Agentic Commerce Protocol so ChatGPT can show richer results with prices, images, reviews, availability, and product details. My move this week: audit the feed, not the blog calendar. If ChatGPT cannot trust your product metadata, your product will lose the comparison before the shopper ever reaches your site.
The shift: the feed is now a discovery asset
ChatGPT shopping is moving closer to structured product retrieval than classic search ranking. OpenAI says its March 2026 product discovery rollout uses ACP to support more complete and current catalog data in ChatGPT, including visual browsing and side-by-side product comparisons. (OpenAI)
That matters because the buyer behavior is different. A shopper is not typing one keyword and scanning ten blue links. They are asking for constraints: budget, use case, style, compatibility, timeline, and tradeoffs. Your product feed has to answer those constraints directly.
This is where Machine Relations shows up operationally: the machine is not just reading your brand page; it is deciding whether your structured evidence deserves to enter the recommendation set.
What I would audit before adding content
The first pass should be feed completeness, freshness, and evidence quality. OpenAI's merchant page says feeds help products appear with accurate, up-to-date information, and its ACP feed spec says OpenAI ingests and indexes product metadata for retrieval and ranking in ChatGPT. (OpenAI merchant page, ACP feed spec)
| Feed layer | What to check | Why it matters in ChatGPT |
|---|---|---|
| Required product fields | Title, description, URL, image, price, availability | Basic eligibility for product cards and comparisons |
| Variant data | Size, color, material, model, bundle, compatibility | Lets ChatGPT match constraint-heavy prompts |
| Proof fields | Reviews, ratings, returns, shipping, warranty | Gives the model comparison evidence beyond price |
| Freshness | Price and inventory sync cadence | Prevents stale recommendations and shopper distrust |
| Crawl access | OAI-SearchBot allowed where relevant | Lets ChatGPT discover and cite supporting product pages |
OpenAI's search product discovery guidance also says merchants should avoid blocking OAI-SearchBot if they want products to be discoverable in ChatGPT search. (OpenAI search guidance)
The operator checklist for this week
Treat this as a revenue systems project, not an SEO task. I would assign one owner across growth, ecommerce, and data, then run this sequence:
- Export the top 100 revenue products and compare feed fields against the live PDP.
- Flag every missing attribute a shopper would use in a natural-language prompt.
- Add proof fields that help an AI compare: reviews, rating count, shipping speed, returns, warranty, availability, and clear use-case language.
- Check whether OpenAI crawlers can access the pages that support the feed.
- Build a weekly feed-health report: missing fields, stale prices, out-of-stock leakage, and products with weak comparison proof.
Do not start by rewriting category copy. Start where ChatGPT is likely to retrieve structured truth: the feed, the PDP, and the evidence attached to both. Then use the AuthorityTech visibility audit to test whether your brand is visible where buyers are already asking.
FAQ
What is ChatGPT product discovery?
ChatGPT product discovery is OpenAI's shopping experience for surfacing products in conversational, visual, and comparison-style results using merchant data, crawler-discovered content, and ACP feed integrations.
Do merchants need a product feed for ChatGPT?
Not always. OpenAI says ChatGPT can crawl sites, but feeds give merchants more control over accurate product details, availability, pricing, and richer comparison data.
Is this SEO or Machine Relations?
It is Machine Relations work. SEO optimizes pages for search engines; Machine Relations makes the structured evidence, entity data, and trust signals clear enough for AI systems to retrieve and recommend.
Additional source context
- Stanford AI Index provides longitudinal evidence on AI adoption, capability shifts, and market behavior. (Stanford AI Index Report, 2026).
- Pew Research Center tracks public and organizational context around artificial intelligence adoption. (Pew Research Center artificial intelligence coverage, 2026).
- Reuters maintains current reporting on artificial intelligence markets, platforms, and policy changes. (Reuters artificial intelligence coverage, 2026).