Product Schema
The product schema is a structured data format (usually as JSON-LD following Schema.org) with which you mark up product information such as name, price, availability and reviews on a web page in a machine-readable way. This lets search engines and AI assistants clearly recognize that it is a product and reliably use the details in answers and results.
Why it matters
Without a product schema, a machine has to guess what stands on your page. Is the number a price, a quantity or a postal code? Is the item available? The product schema answers such questions unambiguously, because it clearly names every detail. For classic search, this produces rich snippets, meaning enhanced results with price, stars and availability directly in the result. For AI assistants like ChatGPT or Perplexity the effect is even more important: they need reliable, cleanly structured facts to name your product correctly in an answer. If the markup is missing, the chance that your offering is mentioned or recommended at all drops.
How it works
Technically, the product schema is usually a JSON-LD block that you build invisibly into the source code of a product page. JSON-LD is a lean data format that notes details as key-value pairs, such as name, brand, price or rating. The basis is the vocabulary of Schema.org, a standard backed by Google, Microsoft and others. You define the type Product and add properties like name, offers (with price and availability) as well as aggregateRating. Search engine crawlers and AI crawlers read out this block and take over the values without having to interpret the visible running text. Important: the marked-up details must correspond exactly to what visitors actually see on the page.
Common mistakes
The classic mistake is a schema that does not match the visible content, such as an old price in the code or invented reviews. This is considered manipulation and can lead to penalties. Just as widespread: incomplete details, when price or availability is missing, so that no rich snippets arise. Outdated data is also a problem, when a product has long been sold out but the schema still reports availability. Check every markup with a validation tool and keep prices and stock status updated automatically. Another stumbling block is duplicate or contradictory markup on the same page, which confuses crawlers and lowers the trust value of your details.
Relation to AI recommendations
AI assistants formulate answers from sources they trust. Structured product data makes your offering such a robust source, because price, brand and rating are not interpreted but read out directly. This increases the citability of your page and thus the likelihood of appearing in a purchase recommendation or a comparison. In the context of Generative Engine Optimization, the product schema is therefore a lever for becoming visible in AI answers. It does not replace good content, but it lowers the risk of misrepresentations and hallucinations about your product. Whoever provides machine-readable facts gives the AI less reason to invent details incorrectly.
Example
A small online bike shop sells a trekking bike for 899 euros. On the product page, the operator adds a product schema with name, brand, price, currency, availability (in stock) and an average rating of 4.6 stars from 38 reviews. If someone asks an AI assistant for a solid trekking bike under 1,000 euros, it can take over the cleanly marked-up values and name the bike along with its price and rating. Without the schema, the AI would have to laboriously guess the price from the running text and, in doubt, would not mention the shop at all.
Common questions
Do I need programming skills for a product schema?
Not necessarily. Many shop systems like Shopify or WooCommerce generate the product schema automatically or via plugins. For custom pages, a JSON-LD generator helps create the finished code block, which you then insert into the page.
How do I check whether my product schema is correct?
Use Google's Rich Results Test or the Schema Markup Validator. Both show you errors, warnings and missing required details. Additionally, check whether the marked-up values like price and availability match exactly with the visible page content.