AI shopping tools skip products when the data behind them is too thin to match or compare. A missing barcode or a vague category is usually the cause. Google's Merchant Center help says products with missing or incorrect GTINs may have limited visibility, and most of these gaps are empty fields you can fill in the Shopify admin.
- A shopping assistant compares products on fields, so an empty field is a fact it cannot use.
- Google says a product that has a GTIN but is submitted without it could be disapproved in Merchant Center.
- The fields to fill first on Shopify are the barcode, the product category, the weight, the variant options and the main image.
- In our June 2026 study, 72% of 216 Shopify storefronts failed Entity Integrity, the metric for complete product and brand data.
Why do AI shopping tools skip some products?
A shopper asks for "a waterproof hiking boot in wide fit under $150". The assistant has to match that sentence against product data. A boot with no width option, no stated material and a category of "Shoes" gives it little to match on. A boot that states all three gets compared. You will not get a notice when a product is left out of a comparison.
Google is plain about the identifier part. Merchant Center's GTIN help says products with missing or incorrect GTINs may have limited visibility. It also says a product that has a GTIN assigned and is submitted without it could be disapproved. OpenAI lists structured metadata such as price and product description among the inputs for ChatGPT shopping, which how ChatGPT shopping works covers.
We score this layer as Entity Integrity. In our June 2026 study of 216 Shopify storefronts, 72% of stores failed it, more than any other of the five areas in that study.
The figure is an illustration. No shopping system publishes its exact rules.
Which product fields matter most?
| Shopify field | Where to find it | What goes wrong when it is empty |
|---|---|---|
| Barcode (GTIN, UPC, EAN, ISBN) | Product, then the variant, under Inventory | The product cannot be matched to the item a shopper means |
| Category | Product page, Category field | The product is grouped with the wrong things, or with nothing |
| Weight | Product page, Shipping section | Shipping cannot be worked out |
| Variant options | Product page, Variants section | Sizes and colors cannot be filtered |
| Main image and alt text | Product page, Media section | The product has no picture, or a picture with no words |
Identifiers deserve their own reading, and GTIN, MPN and SKU explains which one your products need. A handmade product often has no GTIN at all, which is fine. Leave the barcode empty rather than inventing one.
How do I fill these fields in bulk on Shopify?
- In your Shopify admin, go to Products and filter the list to one collection or product type.
- Select all the products in the list and click Bulk edit.
- Click Columns and add the fields you want to check, such as Barcode, Weight, Category and Product type.
- Fill the empty cells. The spreadsheet view shows every variant on its own row, so gaps are easy to spot.
- Click Save, then open one product page on the live store and check the values.
Some specs have no standard field, such as material or heel height. For those, use the category metafields Shopify suggests for the chosen category, or create your own under Settings › Custom data › Products. The product fields AI shopping agents can read explains which ones to create.
Where should a footwear store start?
Take a footwear store on Shopify with 300 products and about 2,000 variants. The boots carry sizes as variants, but width lives only in the description ("runs wide"). The barcodes are empty because the brand never typed them in, though every box from the manufacturer has one printed on it. Half the products have the category left as "Shoes".
I would work in this order. Copy the barcodes from the supplier's sheet into the Barcode column first, because the barcode is the clearest way for a machine to tell one boot from another. Then set the category to the specific type, such as hiking boots. Then add width as a variant option or a metafield, so "wide fit" becomes a fact a machine can filter on. A store this size can do the first two in a day with the bulk editor.
How do I keep the catalog clean after the first pass?
Check new products before they go live, since that is where the gaps come back. Make the barcode, the category and the weight part of your checklist for every new product. Keep the price, stock and variants on the page the same as in your feeds, which live data agents must trust covers.
The Catalog Assessment for up to 1,000 products shows the gaps across a whole catalog at once. It reads your Shopify export and counts how many products carry a description, an SEO title, an SEO description and an image. Image descriptions have their own guide in alt text for product images.
Questions people ask
What is catalog data quality?
Catalog data quality is how complete and accurate the data behind your products is: identifiers, category, weight, variants, specs and images. AI shopping tools compare products on that data. An empty field is a fact the comparison cannot use.
Why is it hard to notice when AI skips my products?
No report tells you. A product that loses a comparison in an AI answer simply does not appear, and no report shows the shortlist it missed. That is why the fields have to be checked on purpose rather than waited on.
Which product fields should I fill first on Shopify?
Start with the barcode for every product that has a GTIN, then the category, the weight and the variant options. Google says products with missing or incorrect GTINs may have limited visibility in Merchant Center. Use Bulk edit under Products to fill them quickly.
What if my products do not have a GTIN?
Leave the barcode empty. Handmade and private label products often have no GTIN, and Google does not ask for one in that case. Never invent a number, since a wrong identifier matches your product to someone else's.
Sources: Google Merchant Center Help, About unique product identifiers, read November 9, 2026; OpenAI, Shopping with ChatGPT Search, read September 26, 2026; Visibility Mesh, The State of AI Visibility 2026, 216 Shopify storefronts scored June 28, 2026. The footwear store is an illustration, not a client.