Glossaire · SEO

Semantic HTML markup

Semantic HTML markup is the practice of using tags whose names describe the role of the content they wrap, rather than its mere appearance. Elements such as <header>, <nav>, <main>, <article>, <section>, <aside> and <footer> explicitly signal the function of each block, while <h1> to <h6> structure headings and <strong>, <em>, <time> or <figure> clarify the meaning of specific content. This structure helps search engines, language models and assistive technologies understand how a page is organized without relying on visual styling. Introduced at scale with HTML5, semantic markup replaces stacks of generic <div> and <span> tags with readable, meaningful code. It forms a technical foundation of SEO: a well-marked-up document is easier to crawl, index, cite and turn into a rich result by search engines and generative AI alike.

Semantic markup means choosing each HTML tag for what it means, not for the visual effect it produces. It is one of the foundational practices of technical SEO, often invisible to the user yet decisive for the machines reading your pages.

How it works

HTML5 introduced a family of structural tags that describe the role of each zone: <header> for the header, <nav> for navigation, <main> for the main content, <article> for a self-contained unit, <aside> for supporting content and <footer> for the page foot. Inside, the heading hierarchy from the H1 tag down to H6 organizes information into logical levels. Other tags such as <time>, <figure>, <figcaption>, <strong> or <em> add meaning to precise fragments.

A crawler or a language model parses this marked-up tree to locate where useful content begins, how it connects and which passages match an intent. Without semantics, a string of <div> elements conveys no functional clue.

Why it matters

Clean markup benefits three audiences. Search engines, which crawl and index more efficiently. Assistive technologies, which rely on structure to convey the page to people with disabilities. And generative AI, which more easily extracts a clean, self-contained passage to cite in an answer.

A retenir

A well-marked-up page is readable by humans and machines alike: it is the prerequisite for solid indexing and strong citability.

A concrete example

Replacing <div class="title"> with a real <h2>, wrapping the article body in <article> and dates in <time datetime="2026-06-15"> is enough to clarify the structure. Combined with structured data in JSON-LD, this semantic base maximizes your chances of appearing as a rich result and being picked up by search engines and AI assistants alike.

FAQ

Questions fréquentes

It is not an isolated, official ranking factor, but it improves crawling, indexing and content understanding. Clean code makes passage extraction easier and helps qualify for rich results, which indirectly supports your positions.

Semantic markup structures the visible HTML through meaningful tags. Structured data, usually in JSON-LD, adds a descriptive layer invisible to the user. The two complement each other to clarify a page's meaning for machines.

Pour aller plus loin