What Are Heading Tags (H1-H6)?
Short answer
Heading tags (H1 through H6) are the HTML elements that structure a page into a hierarchy: one H1 stating what the page is, with H2s and deeper levels organizing its sections. They're the outline machines read first, and in AI retrieval they define the chunks.
Heading Tags (H1-H6) explained
Headings encode a page's structure in the markup itself: the H1 declares the page's subject, H2s mark its major sections, and H3 through H6 nest detail beneath them. Semantically correct usage, hierarchy by meaning, not styled text masquerading as structure, gives every machine reading the page an instant outline: what this page covers, in what pieces, in what relationship.
Their classic SEO weight is real but calibrated: headings help engines understand content and serve as modest relevance signals, with the H1 carrying the clearest topical statement; Google's guidance has long emphasized their structural role over any keyword-stuffing use. Accessibility raises the stakes independently, since screen readers move through pages by heading hierarchy, making sloppy structure a usability failure before it's an SEO one.
AI retrieval promoted headings from outline to architecture. Chunking commonly follows heading boundaries, which makes each heading the doorway to a potentially standalone passage: a descriptive heading followed by a direct, self-contained answer is precisely the unit fan-out retrieval selects, while vague headings ("Learn More", "Our Approach") label chunks no sub-query will ever match. Writing headings as the questions and claims your sections actually answer is now among the cheapest structural optimizations a page can receive. The outline is the interface; write it like one.
In practice
Headings are the first thing I check in any content review, because they reveal whether a page was structured or merely decorated: styled bold text where H2s should be, four H1s, hierarchies that skip levels. My working rule for writers is that the heading outline alone should read as a coherent summary of the page, each heading a question or claim, each section answering it immediately. Do that and you have optimized for scanners, screen readers, snippets, and chunk retrieval in one pass.
Common misconception
People often treat headings as font sizes for visual hierarchy. They're semantic structure machines read by, and in AI retrieval they define chunk boundaries, which makes a vague heading a passage no query can find.
See also
- Machine-Readable ContentMachine-readable content is content structured so that AI systems, crawlers, and agents can parse, extract, and act on it accurately: real HTML text, clear…
- Content ChunkingContent chunking is how retrieval systems split pages into smaller passages, embed each one, and retrieve at the passage level rather than the page level.
- Technical SEOTechnical SEO is the work that lets search engines and AI systems crawl, render, index and understand a website: site architecture, crawler access, page speed…
- Featured SnippetA featured snippet is the highlighted answer box Google places above traditional results, extracting a passage, list, or table directly from a ranking page.
Free email course: Smarter SEO for the Age of AI
Seven short lessons over three weeks: why clicks are falling even when rankings hold, and how to get AI systems to recommend your brand.
Written and reviewed by Tarun Gehani
SEO and AI visibility consultant in Ann Arbor, Michigan, with 17+ years in SEO. He helps B2B SaaS companies, marketing teams and agencies show up in Google and in AI answers.
