Google Structured Data for Ecommerce Product Pages

Google structured data for ecommerce product pages should make the sellable product obvious: what the product is, what it costs, whether it is available, which reviews are visible, and how the page fits into the catalog.

For ecommerce SEO, structured data is not decoration. It is a consistency layer between the visible product page, merchant feeds, reviews, variants, shipping promises, return policies, and the product facts that search systems can safely reuse.

Core schema types for ecommerce

Product and Offer JSON-LD example

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Merino Travel Hoodie",
  "image": ["https://example.com/images/merino-hoodie.jpg"],
  "description": "A lightweight merino wool hoodie for travel and daily wear.",
  "brand": {
    "@type": "Brand",
    "name": "Example Brand"
  },
  "sku": "MH-001",
  "offers": {
    "@type": "Offer",
    "url": "https://example.com/products/merino-travel-hoodie",
    "priceCurrency": "USD",
    "price": "129.00",
    "availability": "https://schema.org/InStock",
    "itemCondition": "https://schema.org/NewCondition"
  }
}

Structured data must match visible content

The most important rule is consistency. If the structured data says InStock but the page says Sold out, the markup becomes less trustworthy. If schema says USD but the shopper sees EUR, your store is creating ambiguity.

How to think about variants

Variant handling is where ecommerce structured data often gets messy. A single product page may include size, color, material, subscription, bundle, or regional pricing differences. Your markup should avoid creating unrelated Product entities that contradict each other.

Reviews and AggregateRating

Review markup is powerful only when it reflects real, visible review content. Review apps can create duplicate or stale markup, especially when themes already output rating data. Audit both the rendered page and JSON-LD sources.

Technical QA case: an empty review form is not a customer rating

A saved WooCommerce demo-page fixture for Bulbasaur contains an empty review state, a rating dropdown and a review textarea. Treating the presence of that widget or its highest selectable value as feedback produced a false review signal. The useful question is whether the page contains actual customer content or a displayed rating, independently of the form and JSON-LD.

<!-- Shortened from the saved QA fixture; not a current live fetch. -->
<p class="woocommerce-noreviews">There are no reviews yet.</p>
<form id="commentform">
  <label for="rating">Your rating</label>
  <select name="rating" id="rating">
    <option value="">Rate…</option>
    <option value="5">Perfect</option>
    <!-- Other input choices omitted. -->
  </select>
  <textarea id="comment" name="comment"></textarea>
</form>

What we fixed in the scanner

We separated review_widget_present, visible_reviews and visible_ratings from has_review_schema and has_aggregate_rating. Form controls, scripts and hidden review content do not establish visible feedback. The parser uses actual review bodies, positive counts or displayed rating evidence, while keeping the structured-data inventory separate. Schema diagnostics, platform diagnostics and the AI advisor receive those same facts.

For the saved empty-form fixture, both review diagnostics are informational and no review repair issue is generated. Fallback JSON-LD does not invent brand, review or aggregateRating. Positive review and rating fixtures still trigger the relevant missing-schema investigation, so suppressing the false positive does not suppress real evidence.

Verification and limits

Our local regression set covers empty forms, real review bodies, positive and zero counts, displayed ratings, hidden content and schema-only ratings. A separate production Free scan of the public Abominable Hoodie demo on October 24, 2026 showed USD 69.00, SKU MH09, no brand and no real review or rating evidence; its schema and platform review checks agreed. The Free AI citation did not invent a brand or rating. These are technical checks, not customer outcomes. Paid JSON-LD was locked and was not verified in that Free session; variation inventory was not independently checked.

BreadcrumbList for ecommerce context

Breadcrumb schema helps explain product hierarchy. It should align with visible breadcrumbs and canonical navigation. For products in multiple collections, choose a stable primary path instead of changing breadcrumbs unpredictably.

Shipping and return data

Shipping and return structured data can help clarify commercial terms. Even when you do not mark up every detail, the visible page should make shipping, returns, and warranty information easy to find.

A practical structured data audit process

Google Rich Results Test vs a product page scanner