Skip to content
AgoraNexo Services for online stores

Español

Product data for Google and artificial intelligence

I reviewed one public product page from each of 10 Spanish online stores. Price and availability appeared on all of them; the barcode, on 2.

7 November 2026

Shopping through artificial intelligence assistants has received a great deal of attention in recent months, although its future remains uncertain. In 2026, OpenAI withdrew direct purchasing inside the chat itself and moved it to connected apps, after limited adoption.

There is, however, a practical matter that does not depend on these trends. Google and other systems need to identify the products each store sells, and part of that information comes from the data included in its product pages. When that data is missing, they have less information about the product.

I have been selling online since 2012 and manage around 2,800 products in my own store. To understand the situation of small retailers, I reviewed one public product page from each of 10 Spanish online stores.

Results of the review

DataPages that included it
Name, image, description, price and availability10 of 10
Store's own SKU9 of 10
Brand7 of 10
Barcode (EAN)2 of 10
Manufacturer part number2 of 10
Dimensions or other attributes1 of 10
Shipping and returns on the product page itself0 of 10

These results come from a small sample and are not a study. Only one page per store was reviewed, with no statistical selection. In addition, a store may send this data to Google through other channels, such as Merchant Center, which cannot be checked from outside.

The basic data appeared on every page reviewed. It is therefore worth asking for concrete evidence whenever someone claims that a store is "invisible to artificial intelligence".

Three points worth checking

1. Barcode visible, but missing from the data

On one of the pages, the EAN appeared on screen but not in the part intended for machine reading. If the data already exists, the fix is usually simple. Google recommends including the GTIN, the global identifier that covers the EAN, whenever the product has one.

2. Technical errors in the data block

On one page, quotation marks in the description invalidated the data block. On another, the price was missing when the page loaded and was only filled in afterwards with JavaScript. Google warns that dynamically generated markup can make its crawls less frequent and less reliable, and recommends including it in the initial HTML.

3. Blocked robots

Two of the stores reviewed blocked certain artificial intelligence robots through their robots.txt file. This may be the right decision, but it should be a deliberate one rather than a forgotten setting.

How to check

  1. Review the product spreadsheet. Check whether there are columns for brand, EAN, manufacturer part number and attributes, and how many rows are left empty.
  2. Check the EANs. The EAN includes a check digit that makes it possible to detect certain typing errors.
  3. Test one product page. Google's Rich Results Test can be used for this.
  4. Review shipping and returns. They do not have to appear on every page. Google recommends a single company-wide policy through Organization markup, and also allows setting them up in Search Console or Merchant Center.
  5. Do not fill in data without confirming it. Some products, such as certain spare parts, may have no barcode. In those cases it is better to leave the field empty than to enter incorrect data.

What these measures achieve

Complete data does not guarantee appearing in ChatGPT or Google, nor an increase in sales. It does provide a tidier catalog that is easier to interpret, both for these systems and for the business itself.

Checked on 7 November 2026 against Google's documentation.

Free sample review

If you wish, you can send your product spreadsheet. A free sample of 20 products comes back, showing the data that is missing for Google and artificial intelligence. No passwords or access to your store are needed, and if any data is missing, it is flagged rather than invented.

After that, for catalogs of up to 1,000 products, organizing the spreadsheet costs €190, and organizing it and matching it against the supplier's price list, €290. VAT not included.

This service prepares the data but does not include publishing the products. For that, the catalog upload service is available.

Request the free sample

Fill in the form and a reply follows within 24-48 working hours. You can also write directly to contacto@agoranexo.es. Replies are in writing, in English or Spanish.

Nothing needs to be attached here: the spreadsheet is requested by email afterwards. Messages arrive in a single mailbox and are used only to reply. This website sets no cookies.