Product Schema

Definition

Product Schema refers to structured data markup based on the Schema.org vocabulary that describes product information in a machine-readable format. When added to a product page’s HTML (typically as JSON-LD), it provides search engines and AI systems with explicit, unambiguous details about the product - name, description, price, availability, brand, reviews, images, SKU, and more.

Schema.org’s Product type has been a foundational element of SEO since Google began using it for rich search results. In the agentic commerce era, it has become even more critical. AI agents parsing product pages rely heavily on Product schema to extract structured data without having to interpret visual layouts or parse unstructured text.

The markup doesn’t change what users see on the page. It adds an invisible layer of machine-readable data that ensures AI systems can accurately identify what you sell, what it costs, whether it’s in stock, and how customers rate it.

Why It Matters

Product Schema has transitioned from an SEO best practice to an agentic commerce necessity:

• AI agent data source. When ChatGPT Shopping, Perplexity, or any other AI system encounters a product page, Product schema is often the first data it extracts. Clean, complete schema means accurate representation in AI recommendations. Missing or incomplete schema means the AI has to guess - and guesses are unreliable.
• Rich results in traditional search. Product schema enables rich snippets in Google search - star ratings, price, availability, and review counts displayed directly in search results. These rich results have higher click-through rates than plain text listings.
• AI Overviews sourcing. Google’s AI Overviews preferentially cite pages with complete structured data. Product pages with thorough schema markup are more likely to appear as sources in AI-generated shopping summaries.
• Cross-platform consistency. Schema.org is a universal standard. The same Product markup is understood by Google, Bing, Perplexity, ChatGPT, and any system built to parse structured data. Implementing it once benefits visibility across all platforms.
• Data accuracy. Schema eliminates ambiguity. A price in HTML could be displayed in various formats ($19.99, 19,99 EUR, “from $19”). The offers.price and offers.priceCurrency fields in Product schema provide an unambiguous, machine-parseable price point.

For merchants, incomplete Product schema is one of the most common and fixable problems in AI visibility. Many stores have basic schema (name and price) but miss fields that AI agents value highly: detailed descriptions, brand, SKU, aggregate ratings, review counts, availability, and product conditions.

How It Works

Product Schema is implemented as JSON-LD embedded in the product page’s HTML:

1. Core properties. Every product page should include: name, description, image (array of URLs), brand, sku, and url. These form the minimum viable product identity for AI consumption.

2. Offer details. The offers property contains pricing and availability: price, priceCurrency, availability (InStock, OutOfStock, PreOrder), seller, and priceValidUntil. AI agents use this to give users current, accurate pricing.

3. Reviews and ratings. aggregateRating (average rating and review count) and individual review objects provide social proof data that AI systems weigh when making recommendations. Products with ratings are favored over those without.

4. Product variants. For products with multiple variants (sizes, colors), each variant should have its own offer with specific pricing and availability. This prevents AI agents from showing incorrect prices for the variant a user is interested in.

5. Additional properties. Fields like material, color, size, weight, gtin (barcode), and mpn (manufacturer part number) provide specificity that helps AI agents match products precisely to user queries. A user asking for “cotton crew neck t-shirt size L in navy” needs these details to get an accurate match.

Most e-commerce platforms generate basic Product schema automatically. Shopify includes name, price, and availability. WooCommerce with Yoast or RankMath adds more fields. PrestaShop has built-in structured data. But the default output is rarely complete - merchants should audit their schema using Google’s Rich Results Test and add missing fields through theme customization, plugins, or metafield configuration.

Related Terms

• Open Graph Protocol - A complementary metadata standard used for social sharing and AI previews
• llms.txt - A site-level standard for AI context, while Product schema works at the page level
• AI Visibility Score - A metric that heavily weighs Product schema completeness
• Shopify MCP - A protocol that exposes Shopify product data to AI agents, complementing on-page schema