Générateur de Schema

Schema.org markup is a standardized vocabulary of types and properties that you embed in a web page to tell search engines and AI assistants exactly what your content means, not just what it says. Google, Bing, and Yahoo co-created Schema.org in 2011 so crawlers could parse entities like Product, Person, Event, and Recipe without guessing from context. In practice, a correctly structured JSON-LD block on a product page lets Google display star ratings, price, and stock status directly in search results — those are called rich results. A FAQPage block can expand accordion questions below your listing, increasing vertical real estate and click-through rate by 20 to 30 percent in documented tests. Beyond Google, AI assistants like ChatGPT with Browsing and Perplexity read Schema to attribute sources and pull structured facts into answers. Without Schema, they quote you vaguely or not at all. The vocabulary covers more than 800 types and 1,400 properties. You do not need them all. Start with the half-dozen that match your content category. A blog needs Article and BreadcrumbList. A shop needs Product with Offer and AggregateRating. A local business needs LocalBusiness with opening hours and geo coordinates. Structured data is one of the highest-leverage technical SEO improvements still available in 2025 because most sites still do not implement it correctly. Use this generator to produce valid JSON-LD on your first attempt.

Prioritize the Schema types that directly match pages you already have, starting with the three that apply to almost every site. Add Organization on the homepage to declare your brand name, logo URL, and social profile links. Add WebSite with a SearchAction to enable Google's sitelinks search box. Add BreadcrumbList on every non-homepage URL to display a hierarchy trail in search snippets. Beyond those three, match the type to the page content precisely. Use Article or BlogPosting on editorial content. Use Product with Offer and AggregateRating on any item you sell, including price, currency, and availability status. Use FAQPage only where the questions and answers are visibly rendered on the page — Google verifies this. Use HowTo on numbered, step-by-step instructional guides. Use Recipe with cookTime, prepTime, nutrition, and recipeYield for cooking pages. Use Event with startDate, endDate, and location for event listings. Use LocalBusiness with addressLocality, telephone, and openingHoursSpecification for physical locations. Use JobPosting with baseSalary and hiringOrganization for job boards. Avoid adding Schema for content that does not appear on the page. Google classifies that as structured data spam and will remove your rich result eligibility algorithmically. Each type carries required properties and recommended properties. The required ones must be present or the schema is invalid. The recommended ones improve your chances of qualifying for a rich result. This generator enforces required fields and flags missing recommended ones as warnings.

Use JSON-LD in every new implementation from 2025 onward. Google's official documentation has stated a preference for JSON-LD since 2017, and that recommendation has grown stronger every year since. JSON-LD lives in a single script block with type application/ld+json. You place it in the head or body of the page. It is completely separate from the visible HTML, which means you can add, edit, or remove structured data without touching the markup that users see. Developers find it significantly easier to read, generate programmatically, and debug than the alternatives. Microdata embeds structured data as HTML attributes alongside the content it describes. It works but forces you to restructure your HTML around schema requirements, which leads to brittle templates and merge conflicts. RDFa offers similar inline embedding with XML-like attributes. It supports richer semantic graphs than Microdata but carries the same structural coupling problem. Both formats are maintained only for legacy compatibility. No new feature in Google's documentation uses them as the primary example anymore. CMS plugins for WordPress, Shopify, and Webflow all output JSON-LD. Headless frameworks like Next.js and Nuxt.js inject it easily through the head component or script tag. Our generator produces JSON-LD exclusively. Copy the output, wrap it in a script tag of type application/ld+json, and paste it into your page. Test with Google's Rich Results Test before deploying.

Run two tools in parallel to get full coverage. First, use Google's Rich Results Test at search.google.com/test/rich-results. Paste your URL or your raw JSON-LD code directly. The test tells you which rich result types your markup qualifies for, which required properties are missing, and which warnings may reduce your eligibility. It also shows a preview of how the rich result would appear in search. Second, use the Schema.org Validator at validator.schema.org. This tool checks structural correctness against the full Schema.org specification, independent of what Google specifically rewards. It catches property type mismatches, incorrect nesting, and vocabulary errors that the Google test may not flag. Use both because they answer different questions. If the Rich Results Test rejects your markup but the Schema.org Validator accepts it, your schema is structurally valid but does not meet Google's stricter eligibility rules. A common cause is adding FAQPage schema to a page where the questions and answers are loaded by JavaScript after page parse — Google does not execute all JavaScript when rendering structured data. Another common cause is using Product schema without a required Offer property. Our generator pre-fills every required field and warns about missing recommended properties. The output passes both validators by default. After deploying, wait two to four weeks and check Google Search Console under the Enhancements section to confirm the rich result is indexed. Practical tip: bookmark both validator URLs and test on every deploy.