Shopify product schema and structured data

The mental model: schema is the metadata layer of your storefront, written in a vocabulary every Google index, Bing index, ChatGPT-User retrieval, Perplexity-User crawl, and Gemini search uses to identify entities. The HTML Shopify renders describes what a human sees. The JSON-LD your theme emits describes what a machine reads. The same product page can be a clean PDP to a customer and a clearly-typed Product entity with an attached Offer, Brand, AggregateRating, and shippingDetails to Google's understanding layer at the same time — and the two layers do not interfere.

 <script type="application/ld+json"> { "@context": "https://schema.org", "@type": "Product", "name": "Your product title" } </script> 

The vocabulary lives at schema.org⁷, version 30.0 (released 2026-03-19). On a typical Shopify store you'll use about fourteen of its types — Product, Offer, Brand, AggregateRating, Review, BreadcrumbList, Organization, WebSite, Article, FAQPage, MerchantReturnPolicy, OfferShippingDetails, plus the GTIN identifier properties (gtin8 / gtin12 / gtin13 / gtin14) inside Product. The library on this pillar covers all of them.

The reason is structural. Google has been parsing pages for two decades and inferring product entities from clean HTML reasonably well; structured data was a nice-to-have unless you were chasing a specific rich result. AI shopping agents, by contrast, are newer and rely on retrieval-time parsing rather than years of accumulated entity graphs. When ChatGPT-User or Perplexity-User fetches a Shopify PDP to consider for citation, it has seconds to decide what the product is, who makes it, what it costs, whether it ships to the customer, and whether the merchant accepts returns — and JSON-LD answers all five questions on a single read.

The implication: layering on a Shopify storefront is mostly additive, not replacement. Author a complementary JSON-LD block that adds the missing entities without re-declaring the fields Shopify already emits. Two top-level Product blocks with overlapping fields produce duplicate-entity warnings in the Rich Results Test; one Shopify auto-emission plus one merchant-authored @graph that references the existing Product by @id is the clean pattern.

This changes install priority on Shopify. Pre-2024 the high-value schema types were Product + Review + FAQ + HowTo. Today the high-value types are Product + Offer + Brand + AggregateRating + GTIN for Shopping/Merchant Center surfaces, plus Organization + WebSite + BreadcrumbList for AI-engine entity attribution. FAQPage still belongs on the right Shopify pages — the FAQ hub itself, a knowledge-base index — but the visible accordion result is gone for general Search.

The split between theme.liquid and section files matters because of cardinality. An Organization entity describes the brand — it should fire once per page, defined once, in theme.liquid where every template inherits it. A Product entity describes one specific product — it has to be defined inside main-product.liquid (or product-template.liquid on older themes) so the Liquid variables resolve to the current product. Putting Product schema in theme.liquid means the variables are unbound on non-product pages and the schema fails to render.

The validators answer different questions. Rich Results Test answers "will Google show this as an enhancement?" (a Google-product question). Schema.org validator answers "is the markup structurally valid?" (a vocabulary question). A Shopify page can pass the Schema.org validator and fail to trigger a Google rich result — that's the FAQ scenario today, and that's fine if your goal is AI engine extraction rather than SERP visibility. Conversely, a page that fails the Schema.org validator probably won't be reliably extracted by anyone, AI engine or Google.

Each leaf below covers one type. The structure is the same on every page: what the schema describes, when to use it, the current 2026 rich-results status from Google Search Central, the required and recommended fields per Schema.org, what Shopify themes auto-emit (so you don't duplicate), a full copy-paste JSON-LD example for a realistic Shopify product/page, where to paste it inside theme.liquid or the section file, validation steps, and the gotchas that catch first-time installs.

Start with the leaf that matches the page you're working on. If you're shipping a fresh Catalog Readiness install on a new Shopify store, the order is: Organization first, WebSite second, Product third, then everything inside Product (Offer, Brand, AggregateRating, GTIN, hasMerchantReturnPolicy, shippingDetails) in a single per-template block, then BreadcrumbList, then Article on the blog if there is one.

Practical rule of thumb. If you sell physical goods on Shopify, ship hasMerchantReturnPolicy referencing your Shopify policy page and shippingDetails describing your real shipping rates and delivery windows. The work is two MerchantReturnPolicy + OfferShippingDetails objects inside main-product.liquid using Liquid variables. The full implementation walkthroughs live on the hasMerchantReturnPolicy leaf and the shippingDetails leaf.

The fixes are quick once you know what to look for. Author the JSON in a plain text editor (VS Code, not the Shopify admin code editor's rich text), use Liquid variables for every product-specific field, view source on the live page after deploying via Theme Editor / ThemeKit / Shopify CLI, and confirm the JSON renders as JSON rather than as escaped HTML. Run the Rich Results Test on one URL per template type. The full diagnostic list lives on the JSON-LD via theme.liquid leaf.