8 Shopify Schema Fields AI Assistants Check Before Recommending You

September 10, 2026

By Steve Merrill, Founder of WRKNG Digital | September 10, 2026

An AI assistant reads your schema before it reads your design. Eight fields decide most of it. Fill them with real values and you're in the running. Leave them blank and you're invisible, no matter how good the page looks.

I've checked this on dozens of Shopify stores. The schema skeleton is usually there. The values are usually missing. Here are the eight fields that matter most.

Which schema fields do AI assistants actually read?

They read the fields that answer a buyer's question and prove the offer is real. Price, availability, and identity come first. Everything below is a field a model checks before it names you. Validate them with the Schema.org validator so you know the values aren't blank.

1. price

The single most-checked field. A model needs a live, exact price to include you in a budget question. No price means you can't answer "under fifty dollars," so you're dropped.

Keep it current and matched to your feed. A stale price reads as an unreliable offer.

2. availability

In stock or not, in machine-readable form. An assistant won't recommend a product it can't confirm is buyable. Out-of-date availability gets you skipped fast.

Tie it to real inventory so the status is true at the moment the agent checks.

3. gtin or mpn

A product identifier lets a model match your item to the wider world of data and reviews. Follow the Google Merchant product data specification for the right ID.

Without an identifier, your product floats alone with no external proof behind it.

4. brand

A named brand ties the product to a reputation a model can weigh. It's a small field that carries trust.

Leave it empty and the product reads as generic, which lowers its odds in a crowded answer.

5. aggregateRating and review

Ratings and review counts, marked up as data, tell the assistant the product is safe to recommend. Use the Google review snippet structured data.

No structured reviews reads as unknown risk. A model avoids unknown risk when it's picking for a buyer.

6. additionalProperty

This is where your real specs live. Material, weight, dimensions, compatibility. The properties that don't fit a standard field go here as named pairs in your Product schema.

This is the field that proves a fit. Miss it and the model can't confirm the product matches the request.

7. image

A clean, direct image URL the model can attach to the product. Assistants increasingly show a picture with a recommendation.

A missing or broken image URL makes your product harder to present, which quietly drops your odds.

8. description

The schema description should carry the quotable, factual summary of the product. Not marketing fluff. The one sentence a model can read as an answer.

When this field states the fact plainly, you give the assistant something clean to cite.

Further reading

Frequently Asked Questions

What schema fields matter most for AI recommendations?

Price and availability first, then product identifiers like GTIN or MPN, brand, aggregateRating and review, additionalProperty for real specs, image, and description. These are the fields an AI assistant checks to confirm the offer is real and the product fits the request.

Why does empty schema hurt AI visibility?

Because a model reads the raw field values, not your design. An empty field reads as no field, so a product with a schema skeleton but blank values looks incomplete and gets skipped in favor of a store with real, validated values.

Where do product specs go in schema?

In additionalProperty as named pairs like material, weight, and dimensions. Standard fields cover price and availability, but the specs that prove a buyer's fit belong in additionalProperty so a model can read them as structured data.

How do I know if my schema values are real?

Run the page through the Schema.org validator. It shows whether each field has a live value or is blank. Most themes ship the schema structure with empty values, so validating confirms the model will actually see real data.

Want to know if AI assistants can actually find and recommend your store? Get a free AI-visibility read on your Shopify store at WRKNG Digital. We show you exactly what ChatGPT, Perplexity, and Google AI see when they look at your products.

Back to Blog