WooCommerce Product Data: The Fields AI Search Needs And Your Store Is Missing

Most WooCommerce stores publish product structured data that is technically valid and commercially useless. It passes validation, it earns no rich results, and it gives an AI shopping assistant almost nothing to work with.

This guide covers exactly what WooCommerce outputs by default, what Google asks for, which fields AI assistants actually read, and how to close the gap without rebuilding your store.

What WooCommerce Gives You Out Of The Box

WooCommerce generates Product JSON-LD automatically. By default that markup carries the product name, description, image, SKU, and an offers block with price, currency and availability.

That is a real foundation, and it is more than many platforms manage. But it stops there. Out of the box WooCommerce does not output a brand, does not output product identifiers such as GTIN or MPN, and does not output aggregateRating or review data even on stores where customer reviews are switched on and visible on the page.

The result is markup that says a thing exists and costs money, and nothing else. No manufacturer. No identifier that ties your listing to the same product elsewhere. No proof anyone has bought it and liked it.

What Google Actually Asks For

Google separates its product documentation into product snippets and merchant listings. If people can buy on your page, merchant listings is the one that applies to you.

The required properties for a merchant listing are short: name, image, and an offers block containing price (or priceSpecification.price) and priceCurrency. WooCommerce already clears that bar, which is why validation passes and store owners assume the job is done.

The recommended list is where the value sits, and it is much longer:

• At the product level: brand.name, description, sku, mpn, and the global identifiers gtin, gtin8, gtin12, gtin13, gtin14 or isbn, plus aggregateRating and review.
• At the offer level: availability, itemCondition, url, priceValidUntil, shippingDetails, and hasMerchantReturnPolicy.

Google describes these as recommended rather than required, and that word does a lot of damage. Recommended here does not mean optional in any practical sense. It means these are the fields that decide whether your product is eligible for the richer displays, whether it can be matched confidently against the same product sold elsewhere, and whether it can be compared on real total cost rather than sticker price.

The Return Policy Rule That Caught Stores Out

Inside the return policy markup there is a genuine requirement that many stores missed.

Google made returnPolicyCountry a required field within MerchantReturnPolicy, using two-letter ISO 3166-1 alpha-2 country codes. The recommended fields alongside it are merchantReturnDays, returnFees, returnMethod and returnShippingFeesAmount.

If you added return policy markup before that change and never revisited it, there is a reasonable chance your markup is now incomplete. Search Console will usually tell you, but only if someone is reading it.

Why AI Assistants Care More Than Google Does

Here is the part most WooCommerce guides skip.

When a shopper asks an assistant to find them a product, the assistant is doing something a search engine never had to do. It is not returning ten links for a human to sort through. It is picking two or three products and standing behind the choice. To do that it needs to answer questions your markup may not let it answer.

Which product is this, exactly? That is what GTIN is for. Without a global identifier your listing is a string of text that may or may not describe the same item another retailer is selling. With one, your product can be matched, compared, and included.

Who makes it? That is brand. An assistant asked for a specific brand cannot recommend you if your markup never names one.

What will it actually cost me? That is shippingDetails. A cheaper product with expensive delivery loses a comparison it appeared to win, and an assistant that cannot see your shipping terms will often skip you rather than guess.

What happens if it is wrong? That is hasMerchantReturnPolicy. Return terms reduce the risk of the recommendation, which matters more when an agent is choosing on someone's behalf.

Is it any good? That is aggregateRating and review. Without rating data there is no confidence signal at all, and an assistant weighing three similar products will favour the two that have one.

We covered the wider shift in agentic commerce and what AI shopping agents mean for Indian brands. This is the product-level version of the same problem, and it is the layer you control completely.

How To Fix It In WooCommerce

The good news is that almost none of this needs custom development.

Start with identifiers, because WooCommerce now has a home for them. WooCommerce 9.2 added a field called global_unique_id to the classic product editor, labelled GTIN, UPC, EAN or ISBN. It behaves like the SKU field, accepts text, enforces uniqueness across products, and works for both products and variations. If your store runs 9.2 or later the field is already sitting in the product data panel waiting to be filled.

Filling it is the tedious part and there is no shortcut. Get the barcodes from your supplier or from the packaging and enter them. For a catalogue of a few hundred products a spreadsheet import is far faster than doing it by hand.

Add brand deliberately. Decide whether your products carry a manufacturer brand or your own, set it consistently, and make sure whatever SEO plugin you use is passing it into the markup rather than leaving it blank.

Set shipping and returns once, properly. Both can be defined at the merchant level rather than repeated per product, which makes this a configuration job rather than a content job. Include returnPolicyCountry, because it is required, and add the return window and who pays return postage.

Turn reviews into data, not decoration. Review stars on the page do nothing for search or AI if they are not in the markup. Make sure your setup outputs aggregateRating and review, and if you have no reviews to output, that is the actual problem to solve first. A review request in your post-purchase email sequence is the cheapest fix on this list.

Check what you are really publishing. Run your own product URLs through Google's Rich Results Test and read the merchant listing warnings rather than only the errors. Errors block eligibility. Warnings are the recommended fields you are leaving empty, and those warnings are the actual to-do list. Search Console's merchant listings report will show the same gaps across your whole catalogue.

Our schema markup guide for Indian businesses covers the underlying concepts, and the free schema generator will produce markup you can compare against what your store is currently emitting.

The Order I Would Do This In

If you are looking at a store with none of this in place, sequence matters more than completeness.

• Fix brand and identifiers first. These are what make your product matchable at all, and everything downstream depends on them.
• Add shipping and returns next. One configuration change covers the whole catalogue, so the effort to impact ratio is the best on the list.
• Start collecting reviews immediately, because this is the only item with a lag. Markup can be added in an afternoon. A body of genuine reviews takes months, so the sooner it starts the better.
• Clean up product titles and descriptions last. It matters, but it matters less than being identifiable, and it is the slowest work.

Mistakes To Avoid

• Treating a passing validation test as a finished job. Passing means you met the required minimum, not that you are competitive.
• Reading warnings as cosmetic. In merchant listings the warnings are the opportunity.
• Adding GTINs that are wrong or invented. A bad identifier is worse than none, because it matches your product to something it is not. If you do not have a real one, leave the field empty.
• Marking up ratings that do not exist on the page. Google requires review markup to reflect genuine reviews visible to users, and fabricated ratings risk a manual action.
• Stacking three SEO plugins that each output their own Product block. Duplicate and conflicting markup is a common WooCommerce problem and it confuses everything reading the page.
• Assuming your marketplace listings cover you. Your listing on a marketplace helps that marketplace. It does nothing for your own store's visibility in AI answers.

Frequently Asked Questions

Does WooCommerce add product schema automatically?

Yes. WooCommerce outputs Product JSON-LD with name, description, image, SKU and an offers block containing price, currency and availability. It does not add brand, GTIN, MPN, or rating and review data by default.

Is a GTIN mandatory?

Google lists global identifiers as recommended rather than required, so a product without one can still appear. In practice an identifier is what allows your product to be confidently matched to the same item elsewhere, which is exactly what comparison and recommendation depend on. Products that genuinely have no GTIN, such as handmade or custom items, are a legitimate exception.

Will this help me rank in normal Google search?

Indirectly. Structured data is not a ranking factor on its own, but it decides eligibility for richer result formats, and those formats change how much of the result page you occupy and how many people click. See our WooCommerce SEO guide for India for the ranking side of the work.

How do I know if AI assistants can currently see my products?

Ask them. Search for your product category the way a customer would phrase it, note whether you are named, and check which sources are cited instead. Then compare your product markup against those sources. Our guide on measuring AI search visibility covers a repeatable version of this.

Do I need a developer for any of it?

Mostly no. Identifiers, brand, shipping and return settings are all configuration and data entry. A developer usually only becomes necessary when plugins are producing duplicate Product blocks that need untangling, or when a custom theme has overridden the default output.

The quickest way to see where your store stands is to run it through our free AEO Checker. If you want your product data, schema and review setup fixed properly rather than audited and handed back, book a free strategy call with Growzai and we will show you exactly what an AI assistant can and cannot see on your store today.