A practical Product schema guide for stores that want better visibility in ChatGPT, Google, and AI-driven product discovery systems.
If ChatGPT does not know your product exists, it will recommend a competitor. That is not an exaggeration. AI systems already shape recommendation lists for millions of users, and if your store is invisible to them, you simply do not exist in that game.
In the previous article on Agentic Commerce I explained how AI agents search, compare, and guide users toward the right products. And in the ChatGPT Shopping guide I covered how recommendations inside ChatGPT work. The key takeaway from both: AI finds products through structured data, not through pretty layout or page copy.
In this article we will check whether AI can actually see your products and fix that in 2-3 hours. No code, and no developer in most cases. Step by step for Shopify and WooCommerce.
Why AI cannot see your products
When you open a product page, you see the photo, the price, and the Buy button. You understand what the product is. But AI does not look at the page like a human. It does not see the design, it reads the code.
More specifically, AI looks for special markup, schema.org/Product. That is a set of structured fields that clearly describe the product in machine language: name, price, currency, availability, and brand.
If that markup is missing or incomplete, this is what happens:
- Google: without markup it only shows the page title and URL. With markup it can show rating stars, price, and availability directly in search results, the so-called rich results.
- ChatGPT: without structured data it either leaves the product out of recommendations or uses incomplete information.
- Google Gemini: the same pattern, products with complete markup get priority.
Google officially says that structured data helps the search engine understand the content of a page and show rich results.
And OpenAIโs ChatGPT Shopping documentation explicitly says that a product must have structured data, including price, availability, and features, to appear in recommendations.
Schema is like the product label in a supermarket. Without it, the buyer only sees the package and has to guess what is inside. With the label, they instantly understand the name, ingredients, price, and expiration date. AI works the same way.
Which fields are critical for AI visibility
Not every field matters equally. Here are the 9 key ones that determine whether AI sees your product and how it presents it:
name
Product name. AI searches by name. If the name is โItem 12345โ instead of โFlex Pro ergonomic office chair,โ AI will simply not find it.
description
Product description. AI uses the description to compare products with competitors. The more concrete the description is, the higher the chance of showing up in recommendations.
price + priceCurrency
Price and currency. This appears in Google rich results and helps ChatGPT filter by the userโs budget.
availability
Availability. InStock, OutOfStock, PreOrder. Google filters on it, and ChatGPT will not recommend products that are out of stock.
brand
Brand. AI groups products by brand. When a user asks for โthe best Nike sneakers,โ your product will not appear without a brand field.
review / aggregateRating
Reviews and rating. The stars in Google search. AI analyzes not only the star count but also the review content when making recommendations.
image
Image. Required for product cards in Google and ChatGPT Shopping. Without an image, the product looks suspicious.
offers
Offers. Detailed purchase information: seller, delivery terms, warranty. Makes the product more understandable for AI.
gtin / sku
Product code. Unique identification. Google uses GTIN to match products across stores and verify authenticity.
The minimum starting set is name + price + availability + image. Add the rest step by step. Every new field improves visibility.
Checking your store
Below are step-by-step instructions for the two most popular platforms. Choose yours:
What Shopify already does automatically
Shopify generates basic Product schema out of the box. If you filled in the product card, the name, price, description, and image already go into the markup automatically. That is good, but not enough.
What you still need to add
- Reviews and rating. Shopify does not add review or aggregateRating markup without a dedicated app. Without reviews in markup, there are no stars in Google and no social proof for ChatGPT.
- GTIN / barcode. In Shopify admin there is a โBarcode (ISBN, UPC, GTIN, etc.)โ field in Inventory. Fill it for every product. It is the unique identifier Google uses to match products.
- Brand. Shopify does not have a standard brand field. Add it via a metafield, Settings โ Custom data โ Products โ Add definition โ Brand, or through an app.
- Detailed availability. Make sure Track quantity is enabled and that Shopify is passing InStock or OutOfStock correctly.
Recommended apps
- JSON-LD for SEO by Ilana Davis – from $14.99/month. Automatically generates full Product schema with all fields.
- Smart SEO – $4.99/month. JSON-LD markup plus meta tags and image alt text.
- Judge.me Product Reviews – free plan. Collects reviews and adds aggregateRating to schema automatically.
Step-by-step instructions
- Install Judge.me or a similar review app. In the settings, turn on Schema markup so the app automatically adds review and aggregateRating to each productโs markup.
- Fill out Barcode for every product. Go to Products โ select a product โ Inventory section โ Barcode field. If you do not have a GTIN, use your internal SKU.
- Add a Brand metafield. Settings โ Custom data โ Products โ Add definition. Name: Brand, type: Single line text. Then fill it for each product.
- Check the result in Google Rich Results Test, described below. Paste the URL of any product and see which fields are present and what is missing.
Need help with schema setup? Message me and I will audit your Shopify store. Details
What Yoast SEO and RankMath already do
If you have Yoast SEO or RankMath installed, the basic Product schema is already generated: name, description, price, image, and availability. Both plugins integrate with WooCommerce automatically.
What they do NOT cover
- aggregateRating. Yoast and RankMath do not add review or rating schema automatically. You need a separate plugin or a manual setup.
- GTIN. WooCommerce does not have a standard field for GTIN. You need a plugin or a custom field.
- Brand. Same story, there is no standard field. RankMath lets you set brand in schema settings, Yoast does not.
- Detailed offers. The basic markup does not include seller info, shipping terms, or warranty.
Recommended plugins
- Schema Pro – $79/year. Full Product schema with all fields, GTIN, brand, and offers. WooCommerce integration.
- WP Review – free. Adds review schema and star ratings for products.
- Schema & Structured Data for WP & AMP – free. A lightweight plugin for adding custom schema fields without code.
Step-by-step instructions
- Check the current state. Open any product on the site, right click โ View source, then search for Product or schema. See which fields are already there.
- Add GTIN and Brand. If you use RankMath, go to RankMath โ Schema โ Product and fill in Brand and GTIN. If you use Yoast, install Schema Pro or add custom fields manually.
- Set up reviews. Enable WooCommerce Reviews, WooCommerce โ Settings โ Products โ Enable reviews, or install WP Review for extended functionality.
- Check the result in Google Rich Results Test. For WooCommerce, it is especially important to verify that plugins do not conflict and do not generate duplicate markup.
Need help with schema setup? Message me and I will audit your store. Details
How to check the markup
You have set it up, now let us check that everything works. There are three ways, and I recommend using all three.
Google Rich Results Test
This is Googleโs official tool for checking structured data. It works like this:
- Go to search.google.com/test/rich-results
- Paste the URL of the product page
- Click โTest URLโ
- After 10-20 seconds you will see the result: which schema types were found, which fields are filled, and where the errors or warnings are
A green checkmark next to Product means the basic markup is in place. Yellow warnings mean recommended fields are missing, such as review or brand. Red errors mean critical problems that block rich results.
Schema.org Validator
This is the deeper check. Go to validator.schema.org and paste the URL. The tool shows the full structure of the markup, all nested objects, data types, and relationships. It is useful for verifying that offers, brand, and aggregateRating are connected correctly.
Manual test in ChatGPT
The most interesting method. Open ChatGPT and write a prompt that should make your product appear:
- โRecommend a [type of product] in [city/country]โ
- โWhat is the best [category] under [budget]?โ
- โCompare [your product] and [competitor]โ
If your product does not appear in the answer, that is a signal to act. If it does appear, but with incomplete information, the markup needs more work.
Do not expect instant results. Google usually reindexes a page 2-4 weeks after markup changes. ChatGPT updates its knowledge base even more slowly. Check again after a month.
How long it takes to see results
One of the most common questions is, โI did everything, when will I see an effect?โ Here are realistic timelines:
- Google Rich Results: 2-4 weeks. After you add or fix schema, Google needs to reindex the page. You can speed that up in Google Search Console โ URL Inspection โ Request Indexing. Most competitors have not even started this work, you are already ahead.
- ChatGPT/Gemini recommendations: 1-3 months. AI systems update their data less frequently than Google. But once your product appears, it tends to stay in recommendations for a long time.
The main thing is not to stop after the first setup. Add reviews, update prices, and fill in new fields. Every improvement is another signal to AI that your store deserves attention.
FAQ
Do I need a developer to set up schema?
You can do the basics yourself with plugins. Installing apps such as Judge.me and filling in fields in the admin panel does not require technical knowledge. But for full optimization it is better to bring in a specialist who can check all fields, remove plugin conflicts, and set the markup up correctly. That is especially important for WooCommerce with a custom configuration.
Shopify already generates schema, so why add anything?
Shopify generates the basic markup: name, price, description, and image. But without reviews, GTIN, brand, and detailed offers, your product looks empty to AI. Google will not show rating stars, and ChatGPT will not be able to compare your product properly with competitors. Basic schema is a foundation, not a finished house.
How do I check whether ChatGPT can see my products?
Open ChatGPT and ask a prompt where your product should appear. For example, โRecommend a [product type] in [city]โ or โWhat is the best [category] under [budget]?โ If your product does not appear in the answer, the schema needs work. Repeat the test in 1-2 months after the changes.
How much does schema markup setup cost?
From $0 to $500, depending on the approach. DIY with free plugins such as Judge.me, WP Review, and Yoast SEO is $0. Paid plugins such as JSON-LD for SEO from $14.99/month, Smart SEO for $4.99/month, or Schema Pro for $79/year run from about $5-80/month. If you hire a specialist for a full audit and setup, expect $200-500 one time. For most Shopify stores, free tools are enough.
Product Schema audit checklist + developer brief template
Check your store in 30 minutes and hand a ready task to your developer.
Get it on Telegram →Ad Astra Marketing helps stores become visible in AI search, from Product schema audits to full structured-data setup for Shopify and WooCommerce.
In the next article in the series: full Shopify integration with the AI ecosystem, the catalog in ChatGPT, Google Shopping, and more.
Comments
Discuss on Telegram