Structured data for AI agents: schema.org, JSON-LD and UCP explained

Why structured data is critical for agentic commerce

An AI agent receiving the query "find me a Bluetooth headphone under $80" won't read the marketing copy of your product pages. It will query structured data, information formatted according to standardized schemas that machines can read and compare reliably.

If your product data is well-structured, the agent can extract in milliseconds: the exact price, stock availability, delivery timelines, return policy, aggregated customer reviews, and relevant technical specifications. If your data is poorly structured or missing, the agent moves on to the next competitor.

This is a fundamental difference from classical SEO: in SEO, you can compensate for missing structured data with good text content. In agentic commerce, structured data is not optional, it's the condition for existing in the agentic channel.

Essential schema.org types for agentic e-commerce

1. Product schema

The Product type is the base schema for any product page. For agentic commerce, the minimum required properties are:

name, product name (precise, descriptive, not marketing-speak)
description, factual description of main features
sku or gtin13, unique product identifier
brand, product brand (Brand object)
image, main image URL (high resolution)
offers, price and availability (Offer object)
aggregateRating, average rating and review count

Recommended properties for maximum UCP compatibility:

weight, height, width, depth, dimensions for shipping cost calculation
color, size, for product variants
category, standardized category (prefer standard taxonomies)
additionalProperty, product-type-specific attributes (e.g. Bluetooth 5.3, 30h battery)

2. Offer schema

The Offer object is nested inside Product and defines the sale conditions. This is the information AI agents use most intensively to compare and select:

price, exact price including tax (number, not text)
priceCurrency, currency ("USD", "EUR")
availability, availability ("InStock", "OutOfStock", "PreOrder")
shippingDetails, delivery time and cost (OfferShippingDetails object)
hasMerchantReturnPolicy, link to return policy
seller, merchant information (Organization object)

3. MerchantReturnPolicy schema

Since the UCP deployment in January 2026, MerchantReturnPolicy has become mandatory for compliant merchants. AI agents actively use it to compare return conditions between merchants:

returnPolicyCategory, policy type ("MerchantReturnFiniteReturnWindow", etc.)
merchantReturnDays, number of days to return an item
returnMethod, return method ("ReturnByMail", "ReturnInStore")
refundType, refund type ("FullRefund", "StoreCreditRefund")

4. OfferShippingDetails schema

Delivery times and costs are key factors in agentic purchase decisions, often as important as price. The OfferShippingDetails object must include:

shippingRate, shipping cost (MonetaryAmount object)
deliveryTime, estimated delivery time (ShippingDeliveryTime object)
shippingDestination, geographic delivery zones covered

Complete JSON-LD example for a UCP-compliant product page

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Bluetooth Pro 500 Headphones",
  "sku": "BHP-PRO-500-BLK",
  "gtin13": "0123456789012",
  "description": "Bluetooth 5.3 headphones, active noise cancelling, 30h battery",
  "brand": { "@type": "Brand", "name": "AudioPro" },
  "aggregateRating": {
    "@type": "AggregateRating",
    "ratingValue": 4.6,
    "reviewCount": 284
  },
  "offers": {
    "@type": "Offer",
    "price": 79.99,
    "priceCurrency": "USD",
    "availability": "https://schema.org/InStock",
    "hasMerchantReturnPolicy": {
      "@type": "MerchantReturnPolicy",
      "returnPolicyCategory": "https://schema.org/MerchantReturnFiniteReturnWindow",
      "merchantReturnDays": 30,
      "returnMethod": "https://schema.org/ReturnByMail",
      "refundType": "https://schema.org/FullRefund"
    },
    "shippingDetails": {
      "@type": "OfferShippingDetails",
      "shippingRate": {
        "@type": "MonetaryAmount",
        "value": 0,
        "currency": "USD"
      },
      "deliveryTime": {
        "@type": "ShippingDeliveryTime",
        "handlingTime": { "@type": "QuantitativeValue", "minValue": 0, "maxValue": 1, "unitCode": "DAY" },
        "transitTime": { "@type": "QuantitativeValue", "minValue": 2, "maxValue": 3, "unitCode": "DAY" }
      }
    }
  }
}

How to validate your structured data

Google Rich Results Test (search.google.com/test/rich-results), syntax and rich results eligibility
Schema.org Validator (validator.schema.org), complete JSON-LD validation
UCP Compliance Checker, official tool on the UCP GitHub for full compliance validation
Bing Markup Validator, Bing Copilot compatibility

Structured data for AI agents: the complete schema.org and UCP guide