By Steve Merrill · October 26, 2026
The eight structured data fields that decide whether AI search can read your Shopify products are: Product type and name, Offer price and currency, availability, product identifiers (gtin, mpn, sku), brand, ratings and reviews, image, and shipping plus return policy. Get those right in your JSON-LD and AI shopping engines can parse, trust, and recommend your catalog. Miss them and you're invisible.
I spent 15 years running a $10M clothing brand. I know what it feels like to do everything right on the storefront and still lose the sale somewhere you can't see. That's AI search right now. The buyer never lands on your page. The machine reads your data and makes the call.
Most Shopify themes ship partial markup. They give you a Product type and nothing else worth reading. That gap is why good stores get skipped.
Here's the field-by-field checklist. Validate every one.
1. Product @type and name
Every product page needs a schema.org Product object with a clear name property that matches what's on the page. This is the anchor. AI engines use the Product type to know "this is a thing you can buy" and the name to match it against a shopper's query. No Product type, no product in the answer.
2. Offer price and priceCurrency
Wrap pricing in an Offer with both price and priceCurrency (use ISO 4217 codes like USD). AI shopping tools filter and compare on price, so a missing currency makes your number useless to them. Google's product structured data docs list price and currency as required for merchant listings.
3. availability (ItemAvailability)
Set the Offer's availability to a real ItemAvailability value like InStock or OutOfStock. AI engines will not recommend something they can't confirm is buyable today. Shopify theme stock changes should update this automatically, so test it on an out-of-stock variant and watch what the JSON-LD says.
4. gtin, mpn, and sku
Product identifiers are how machines match your listing to the same product everywhere else. Provide gtin (the barcode) when you have it, plus mpn and sku. Google treats unique product identifiers as a trust signal, and products with them get matched to richer data across the web. Handmade with no barcode? Skip gtin and use mpn and sku.
5. brand
Add a brand property as a Brand object with your name. AI answers group and rank by brand constantly ("best running shoes from X"). If your brand field is blank, you don't exist in those comparisons. That's a free field to fill, and most stores leave it empty. Thirty seconds of work for a spot in the ranking.
6. aggregateRating and review
Include aggregateRating with ratingValue and reviewCount, plus individual review objects when you have them. Social proof is one of the strongest things AI pulls into a recommendation, because it's answering "is this good?" not just "what is this?" Only mark up reviews that actually appear on the page. Faked ratings get you a manual penalty from Google.
7. image
Every Product needs an image property pointing to a high-resolution, directly accessible URL. AI shopping surfaces are visual, and a product with no valid image in its structured data gets dropped from carousels and visual answers. Use absolute URLs, not relative paths, so the crawler can actually fetch the file. One broken image path and you're text-only in a visual feed.
8. shippingDetails and hasMerchantReturnPolicy
Add shippingDetails and hasMerchantReturnPolicy to your Offer. These are the two fields almost every Shopify store misses, and they're exactly what a buyer asks AI before they commit: "how much is shipping and can I return it?" Answer it in the data and you win the ones who are ready to buy.
How We Chose This List
These eight fields come straight from the required and recommended properties in Google's product structured data documentation and the schema.org Product and Offer specs, cross-checked against what AI shopping engines actually parse. We prioritized the fields that make or break whether a product can be recommended at all, not nice-to-haves. Every one of these is something I'd check before I'd spend a dollar driving traffic to the page.
FAQ
Q: Does Shopify add Product structured data automatically?
Some themes inject basic Product JSON-LD. Most ship thin markup that skips gtin, shippingDetails, and return policy. Audit the raw JSON-LD yourself and fill the gaps.
Q: What is the single most important field for AI search?
Offer price with priceCurrency, plus a valid availability value. AI engines won't recommend a product they can't price or confirm is in stock.
Q: Do I need GTINs if my products are handmade?
No. If there's no manufacturer barcode, use mpn and sku instead. Google says to provide the identifiers you actually have, not to invent them.
Q: How do I check my structured data is valid?
Run every product URL through Google's Rich Results Test and the schema.org validator. Both flag missing required fields and broken JSON-LD.
Want this done right across your whole catalog instead of one product at a time? That's the work we do at WRKNG Digital. See how we get Shopify stores ready for agentic commerce.
