Why AI shopping agents skip Shopify products that are published to your Online Store
Your product is published to the Online Store. It has a product page. You can buy it yourself.
So why might an AI shopping agent not show it?
Because publishing to the Online Store is one requirement, not the whole test.
Shopify Catalog has its own rules for stores and products. When a store and product meet them, Shopify says eligible products can become discoverable through AI channels including ChatGPT and Microsoft Copilot. A published product that misses one rule may not make it into that catalog.
That distinction matters when you are checking what shoppers can find through AI.
Start with the product checks
A published product needs a few basic things before Shopify Catalog can include it:
- A product title
- At least one product image
- A price above $0
- A product URL Shopify can identify
- Publication to Online Store, Hydrogen, or Headless
These are worth checking first because they are specific and quick to fix.
Open a product in Shopify now. Check its title, price, and media. Then open the product page in an incognito browser window. If the page does not load for a visitor, an agent cannot use it as a normal product listing either.
A $0 product is an easy one to miss. It may be intentional for a sample, deposit, or item still being prepared. But it does not meet Shopify Catalog's price requirement.
The same applies to products without an image. A product can be technically published while still lacking data Shopify expects for catalog discovery.
Published is not the same as searchable
Two Shopify settings can exclude a product from Shopify Catalog even if it is published:
- Unlisted product status
- The `seo.hidden` metafield
Shopify says Unlisted products are also removed from sitemaps, internet search, storefront search, collections, and recommendations. That status has a clear purpose. You may use it for a private replacement link, a wholesale-only item, or a product not ready for normal discovery.
But it is not a good fit for a product you want AI shopping tools to find.
The `seo.hidden` metafield does something similar for search visibility. A product hidden from search engines is excluded from Shopify Catalog.
This is a useful under-a-minute check:
1. Open Products in Shopify. 2. Choose a product that should be discoverable. 3. Check that its status is not Unlisted. 4. Review its search engine listing and any SEO metafields your theme or apps expose.
Do not remove a setting blindly. Some products should stay hidden. The point is to make sure the setting matches the job of the product.
Your store can block every otherwise-valid product
Sometimes the product is fine. The store is the issue.
Shopify's Catalog requirements also apply at store level. Its guidance lists several things that can prevent eligibility, including:
- An unsupported Shopify plan
- A password-protected or private store
- A B2B-only store
- A locked or restricted store
- A store in development mode
- Account standing or policy issues
Shopify also considers account history. Its requirements include a verified email, two-step authentication, identity or business verification where required, genuine sales, prompt fulfillment, low chargebacks, paid Shopify invoices, and resolved policy issues.
That can feel separate from merchandising. It is still part of whether products can be included.
If a whole catalog appears to be missing, do not spend the afternoon rewriting every description. First check whether the storefront is behind a password page and whether there is an account issue Shopify needs you to resolve.
Eligibility can also change. A product that was once available through an AI channel is not guaranteed to stay there if the store no longer meets Shopify's requirements.
Each AI destination does not use the same route
It is tempting to think of “AI shopping” as one channel. It is not.
ChatGPT and Microsoft Copilot use Shopify Catalog. So the product and store requirements above are directly relevant to them.
Google AI Mode and Gemini use a different connection. Shopify says they require the Google & YouTube sales channel, with product and shipping data synchronized.
Meta uses the Facebook and Instagram by Meta sales channel, with product-feed syncing.
This means a product can be ready for one destination and absent from another.
Say your store sells a backpack. It may meet Shopify Catalog requirements and be available to ChatGPT or Copilot. That does not confirm that the Google & YouTube channel has its product and shipping data synced for Gemini.
When you are diagnosing a missing product, name the destination first. Then check the connection that destination actually uses.
Keep price and inventory current
Agents need current product facts. A shopper who sees an old price or an unavailable item has not had a useful experience.
Shopify says merchants must keep dynamic data, including price and inventory, up to date. For catalogs that change often, Shopify recommends real-time syncing through the GraphQL Admin API or webhooks.
Most merchants do not need to think about APIs every day. The practical question is simpler: when you change a price or stock level, does every sales channel receive that change?
If you use another system for inventory, check its sync status after a recent product update. Pick one product whose price changed this week. Compare the Shopify product record with the product data in the relevant sales channel.
A correct title and image do not help much if the rest of the listing is stale.
What Descriva checks on the product
Descriva's Agent view shows what AI shopping agents get from Shopify Catalog for each product.
It also flags the catalog requirements a product misses, including:
- A draft or Unlisted product
- No image
- A price of $0
- Not being published to the Online Store
- Being hidden from search engines
- A store behind its password page
That gives you a product-by-product place to start, rather than trying to infer catalog eligibility from the storefront alone.
When a product has no category, Descriva can suggest one. It only applies that category after you approve it.
The app does not treat this as a score. AI readiness is Descriva's only score, and it covers five broader areas across the store. Agent view is a checklist of what Shopify Catalog can receive and what is missing for that specific product.
Check the right problem before changing the copy
Product copy still matters. Clear descriptions, specifications, image alt text, and FAQs give shoppers and assistants better information to work with.
But copy cannot fix a product that is Unlisted, priced at $0, hidden from search engines, or blocked by a password page.
Work in this order:
1. Confirm the product meets Shopify Catalog's basic requirements. 2. Check that it is not Unlisted or hidden from search engines. 3. Check store-level eligibility and password protection. 4. Confirm the right sales channel is connected for the AI destination you care about. 5. Make sure price and inventory stay current. 6. Then improve the product information shoppers will read.
That order prevents a common waste of time: polishing a listing that an AI shopping agent cannot receive in the first place.
If you want to see the product-level gaps Shopify Catalog can see, install Descriva and open Agent view for one of your key products.