Product schema is the structured data that lets search engines and AI assistants read your product's name, price, availability, and reviews as facts instead of guessing from page text. If you sell anything online and you want your items to appear in Google's shopping experiences, in AI Overviews, or when someone asks ChatGPT "what's a good waterproof hiking boot under 150 euros," Product markup is what makes your catalog legible to the machines doing the recommending.
What Product schema actually does
An AI shopping answer is only as good as the data it can trust. When an assistant compares products, it needs a reliable source for each item's price, stock status, rating, and key attributes. Your beautifully designed product page communicates all of that to a human, but a model parsing raw HTML can easily miss a price rendered by JavaScript or confuse a "was" price with the current one. Product JSON-LD removes the ambiguity by stating each fact explicitly.
The core type is Product, usually paired with two nested types: Offer, which carries price, currency, and availability, and AggregateRating or Review, which carries your star ratings. If you are new to structured data generally, our guide to JSON-LD and Schema.org covers the syntax and where the markup lives on the page.
The fields that get your products cited
Prioritise the properties that answer buying questions:
- name and description — clear, specific, matching the visible page.
- image — high-quality URLs; assistants and shopping surfaces lean heavily on these.
- offers — price and priceCurrency (use EUR for European stores), plus availability like InStock or OutOfStock, and priceValidUntil for time-limited deals.
- aggregateRating — ratingValue and reviewCount, but only if genuine reviews exist on the page.
- brand, sku, and gtin — identifiers that let systems match your listing to the same product elsewhere and trust it.
Two rules keep you out of trouble. First, never mark up data that is not visible to shoppers on the page — fabricated ratings or hidden prices violate guidelines and can get your rich results suppressed. Second, keep availability accurate in real time; an assistant recommending an out-of-stock item helps no one and trains it to distrust your feed.
Images and speed still decide the sale
Product markup gets you into the answer, but the click still depends on the page delivering. Shopping surfaces and AI results favour products with fast, clean pages and crisp imagery. Compressing and correctly sizing your product photos matters both for the visual result and for load speed — our guide to image optimization with WebP and AVIF walks through doing this without wrecking quality. A product that loads instantly and looks sharp converts better once the assistant has sent the visitor your way.
Getting into AI shopping answers
Structured, trustworthy product data is exactly the kind of content AI Overviews and answer engines like to pull, because it is unambiguous and easy to attribute. The same habits that win there — direct facts, valid markup, consistent identifiers — are what our guide on appearing in Google AI Overviews recommends across the board. Treat every product page as a small, self-contained answer to "should I buy this?" and give the machine the facts it needs to say yes on your behalf.
Validate your markup after every template change, watch for JavaScript that hides the price from crawlers, and keep your offers and stock in sync with your real inventory. Small feed errors compound fast across a large catalog.
Not sure whether crawlers can read your product data as you intended? Run the CheckMy.site scanner to confirm your Product schema is valid, your prices are visible to bots, and your catalog is ready for AI shopping results.