WooCommerce
WooCommerce product feed for AI agents: what to include
A product feed lets AI agents read your whole catalog at once, with current price and stock. Here is what an AI feed contains, why it must stay in sync, and where it goes.
By ShopToAI Team · Updated Jul 16, 2026
Structured data on your product pages is one way an assistant reads you. A product feed is the other, and for AI shopping it is becoming just as important. A feed is a single, structured file of your whole catalog that platforms pull on a schedule, so an agent does not have to crawl every page to know your current price and stock.
On-page schema and a feed do different jobs
They are not either-or. You want both.
- On-page JSON-LD is read when an assistant looks at a specific product page.
- A feed is pulled in bulk, kept current, and used to index and compare your whole catalog at once. It is also what powers direct checkout inside some assistants.
If your schema is clean but you have no feed, you are relying on each page being crawled at the right moment. A feed removes that gamble.
What an AI product feed contains
The fields mirror what assistants need to trust and place a product. OpenAI's product feed for ChatGPT, for example, requires the essentials on every item:
- item_id: a unique identifier for the product.
- title and description: the name and full details.
- brand: the manufacturer or brand name.
- price: the current price with its currency.
- availability: a fixed value like in_stock, out_of_stock, pre_order or backorder.
- image_url and url: the main image and the link to the product page.
- seller_name and seller_url: who is selling and where.
Return policy, shipping and review data are recommended on top, and they help, especially where checkout happens inside the assistant. Google's Merchant feed asks for the same shape of data, so a good feed serves both channels.
The part everyone gets wrong: staleness
A feed is only useful if it is true. A feed that still says in_stock for a product you sold out of last week trains the assistant to distrust you, and it can send a shopper to a dead end. The feed has to stay in sync with your catalog as prices and stock change, and it has to agree with the JSON-LD on the page. Contradictions between the two are a red flag an assistant can see.
Where the feed goes
Two concrete channels today:
- Google Merchant feeds, which feed Google's shopping and AI surfaces.
- OpenAI's merchant feed for ChatGPT, which lets ChatGPT index your products and, where enabled, check out directly.
More assistants are moving the same way, and they read the same kind of structured feed.
How this works on WooCommerce
Woo can export product data, but keeping a feed correct, complete and continuously in sync across a changing catalog is the hard part, and it has to match your on-page schema. shoptoai generates the AI-ready feed and keeps it synced as your catalog changes, alongside the on-page JSON-LD, so the two never disagree. Check your store's readiness with a free audit, see the WooCommerce guide for the full setup, and pricing for plans.
Common questions
If my product pages have JSON-LD, do I still need a feed?
Yes. JSON-LD is read per page when crawled; a feed is pulled in bulk, stays current, and powers indexing and checkout across your whole catalog. They complement each other.
What is the most common feed mistake?
Staleness. A feed that shows the wrong price or in_stock for a sold-out item trains assistants to distrust you and sends shoppers to dead ends. It must stay in sync and agree with the page.