What does Google say about SEO? /
Quick SEO Quiz

Test your SEO knowledge in 5 questions

Less than a minute. Find out how much you really know about Google search.

🕒 ~1 min 🎯 5 questions

Official statement

We recommend the use of JSON-LD for structured data markup, although other types of markup can also be used.
7:20
🎥 Source video

Extracted from a Google Search Central video

⏱ 1h00 💬 EN 📅 14/12/2017 ✂ 10 statements
Watch on YouTube (7:20) →
Other statements from this video 9
  1. 6:17 Pourquoi vos pages techniquement parfaites n'apparaissent-elles pas dans Google ?
  2. 7:54 Faut-il vraiment mettre à jour son sitemap offres d'emploi régulièrement pour ranker ?
  3. 9:20 Pourquoi les erreurs 503 peuvent-elles détruire votre crawl budget ?
  4. 12:52 Comment Google affiche-t-il désormais les avis et salaires dans les résultats d'emploi ?
  5. 19:32 Le balisage d'offres d'emploi sans données de localisation : valide ou pas ?
  6. 23:45 Pourquoi Google pénalise-t-il le balisage structuré sur vos pages de résultats internes ?
  7. 30:06 Que risquez-vous vraiment si Google détecte un abus de balisage structuré sur votre site ?
  8. 44:12 Pourquoi le balisage schema emploi ne garantit-il pas votre positionnement dans les résultats ?
  9. 49:47 Faut-il vraiment enrichir ses données structurées avec tous les champs disponibles ?
📅
Official statement from (8 years ago)
TL;DR

Google clearly favors JSON-LD for structured data while keeping other formats (Microdata, RDFa) open for use. This technical preference is due to the separation of markup and HTML. For an SEO practitioner, this means that by 2025, betting on JSON-LD remains the safest choice to ensure optimal interpretation by crawlers, even though other formats still work.

What you need to understand

What is JSON-LD and why does Google prefer it?

JSON-LD (JavaScript Object Notation for Linked Data) is a structured data markup format that is inserted into the source code of a web page, typically within the <head> tag. Unlike Microdata or RDFa, it is not directly embedded in visible HTML.

Google recommends this format because it separates markup from content. This makes it easier for crawlers that do not need to parse the entire DOM to extract data. For a developer or an SEO, it's also easier to maintain: one can modify the markup without touching the HTML, and vice versa.

Are other formats still accepted by Google?

Yes, Microdata and RDFa remain functional and recognized by Google. The statement 'other types of markup can also be used' confirms that these formats are not deprecated.

But be cautious: recommending JSON-LD means that this is the format that receives the most attention during Google's internal tests. Edge cases, bugs, or new features are first validated on JSON-LD. If a site uses Microdata or RDFa and faces an interpretation problem, it might find itself in a gray support area.

What types of markup does this recommendation cover?

This statement applies to all types of structured data: articles, products, reviews, FAQs, recipes, events, job postings, breadcrumbs, videos, organizations, people, etc. Google values JSON-LD for each of these schemas.

Even for simple markups like breadcrumbs or a company logo, JSON-LD remains the recommended format. This creates consistency across the entire site and avoids mixing formats that can complicate maintenance.

  • JSON-LD is the format favored by Google for all structured data, without exception.
  • Microdata and RDFa are still accepted, but receive less attention in engine developments.
  • The code/content separation is the main technical argument in favor of JSON-LD.
  • No format is officially deprecated, but the recommendation has been clear and consistent for several years.
  • Google's validation tools (Rich Results Test, Search Console) handle all formats, but JSON-LD is the best documented.

SEO Expert opinion

Is this recommendation consistent with field observations?

Absolutely. Since around 2016, JSON-LD has surpassed Microdata in Google's official examples. The documentation, tutorials, and case studies published by Google almost exclusively use this format.

In practice, sites that migrate from Microdata to JSON-LD generally experience no loss in performance, and sometimes even an improvement in the indexing speed of rich snippets. Conversely, a site starting with JSON-LD rarely needs to change format later. It's a sustainable choice.

What nuances should be added to this statement?

Google says 'we recommend', not 'we require'. This means that technically, all formats are equal. No format provides a measurable direct SEO advantage in rankings. It's a matter of practicality, not ranking.

However, some CMS or plugins generate Microdata by default (WordPress with certain themes, for example). Migrating to JSON-LD may represent a considerable development effort. If the existing Microdata markup works well and is validated by Google, there is no urgency to switch. [To verify] if the site experiences recurring interpretation issues, in which case JSON-LD may be a solution.

In what situations could JSON-LD cause issues?

If the site uses client-side rendering (JavaScript) and JSON-LD is dynamically injected after the initial load, one must ensure that Googlebot executes the JavaScript correctly. Otherwise, the markup will not be seen.

Another case: some developers place JSON-LD at the end of the <body> instead of the <head>. It works, but it's less clean and can cause delays if the crawler parses the page from top to bottom. It's better to place the JSON-LD script in the <head>, just before the closing tag, to ensure quick reading.

Warning: JSON-LD must be valid JSON. An extra comma, a missing quotation mark, and the entire markup is ignored. Always validate with a JSON linter before deploying.

Practical impact and recommendations

What practical steps should be taken on an existing site?

If the site already uses JSON-LD, there's nothing to change. Just check that the markup is up to date and complies with Schema.org specs. Google Search Console and the Rich Results Test are your best allies for identifying errors.

If the site uses Microdata or RDFa and is functioning well (rich snippets displayed, no errors in Search Console), there is no urgency. However, if a redesign or technical evolution is planned, this is the right time to migrate to JSON-LD. This will simplify future maintenance.

What mistakes should be avoided when implementing JSON-LD?

The first classic mistake is duplicating markup. If the site already has Microdata in the HTML and JSON-LD is added, it may result in two identical markups for the same entity. Google may interpret this as duplicate or contradictory content.

The second mistake: markup not conforming to the actual content of the page. JSON-LD is invisible to the user, which makes it easy to manipulate. But if the markup describes a product priced at €10 when the page shows €50, Google considers this as spam and may penalize the rich results.

How to ensure the site is compliant?

Use Google’s Rich Results Test for each key page type (product page, article, category page, etc.). The tool detects syntax errors and missing properties.

Then, monitor the Search Console, in the 'Enhancements' section. Google reports markup errors detected during crawling. A spike in errors after a deployment signals a problem to be corrected quickly.

  • Prioritize JSON-LD for any new markup or technical redesign.
  • Place the JSON-LD script in the <head> to optimize reading by crawlers.
  • Validate the JSON with a linter before deploying to production.
  • Avoid duplication between JSON-LD and Microdata/RDFa on the same page.
  • Strictly align the markup with visible content to avoid spam penalties.
  • Test each page type with Google’s Rich Results Test.
Migrating to JSON-LD and establishing robust structured markup requires sharp technical expertise and an in-depth understanding of Google's evolutions. If these optimizations seem complex to manage internally, it may be wise to seek help from a specialized SEO agency for personalized support and to avoid costly mistakes.

❓ Frequently Asked Questions

JSON-LD améliore-t-il le classement SEO d'une page ?
Non, JSON-LD n'est pas un facteur de ranking direct. Il permet d'obtenir des rich snippets, qui peuvent améliorer le CTR, mais Google ne classe pas mieux une page simplement parce qu'elle utilise JSON-LD plutôt que Microdata.
Peut-on mélanger JSON-LD et Microdata sur un même site ?
Techniquement oui, mais c'est déconseillé. Mieux vaut choisir un format et s'y tenir pour éviter les conflits et simplifier la maintenance. Si une migration est en cours, documenter clairement quelle partie du site utilise quel format.
Faut-il absolument migrer de Microdata vers JSON-LD si tout fonctionne ?
Non, pas d'urgence. Si le balisage Microdata est valide, maintenu et génère des rich results, il n'y a pas de raison de tout refaire. Par contre, anticiper une migration lors de la prochaine refonte ou évolution technique.
JSON-LD fonctionne-t-il sur des sites full JavaScript (React, Vue, Angular) ?
Oui, à condition que Googlebot exécute bien le JavaScript et que le JSON-LD soit présent dans le DOM final. Vérifier avec le test d'URL en Search Console pour s'assurer que le balisage est bien crawlé.
Quels types de Schema.org sont les plus impactants pour un site e-commerce ?
Product, AggregateRating, Offer et Breadcrumb sont les schémas prioritaires. Ils permettent d'afficher prix, disponibilité, notes et fil d'Ariane dans les SERPs, ce qui booste le CTR et la visibilité.
🏷 Related Topics
Domain Age & History Structured Data JavaScript & Technical SEO

🎥 From the same video 9

Other SEO insights extracted from this same Google Search Central video · duration 1h00 · published on 14/12/2017

🎥 Watch the full video on YouTube →

Related statements

💬 Comments (0)

Be the first to comment.

2000 characters remaining
🔔

Get real-time analysis of the latest Google SEO declarations

Be the first to know every time a new official Google statement drops — with full expert analysis.

No spam. Unsubscribe in one click.