Header Tags (H1, H2, H3)
Header Tags (H1, H2, H3) are HTML elements used to define the headings and subheadings of a web page - H1 for the main page title, H2 for major sections, H3 for subsections, and H4-H6 for deeper nesting when needed. Both human readers and search engines use header structure to understand what a page is about and how it’s organised.
Foundational on-page SEO and accessibility. Header structure that mirrors content hierarchy serves both - readers scan it for orientation, screen readers use it to navigate, and search engines use it as one of many signals about page subject and organisation.
What each level is actually for
H1. The main title of the page. One per page. Should match (or closely match) the page’s primary topic - what someone would call this page if asked. Different from the page’s `
H2. Major sections of the page. A typical 1,500-word article has 4-7 H2s breaking it into navigable chunks. Each H2 should describe what the section covers cleanly enough that a skim-reader can decide whether to read it.
H3. Subsections within an H2. Useful when an H2 section has multiple sub-points worth their own headings. Common in tutorials (“Step 1,” “Step 2,” “Step 3” as H3s under “How to do X” as H2).
H4-H6. Deeper nesting. Rarely needed in marketing content. Common in technical documentation where deep hierarchies are useful.
The two non-negotiable rules
One H1 per page. Multiple H1s confuse search engines about what the page’s primary topic is. Some CMS templates produce sneaky multiple H1s (logo image as H1, sidebar widget titles as H1, etc.) - worth auditing.
Don’t skip levels. H1 → H2 → H3 → H4 in order. Going from H1 directly to H3 without an H2 between breaks document outline conventions and can confuse screen readers and (less importantly) crawlers.
Where header tag use goes wrong
Three patterns:
Headers used for visual styling rather than structure. “I want this text to be bigger, so I’ll wrap it in H2” - even though it’s not actually a section heading. Mismatches outline structure and visual presentation.
Generic header phrasing. H2s like “Introduction,” “Background,” “Conclusion” tell readers and search engines almost nothing. “What goes wrong with X” or “The three things that matter for Y” are descriptive and improve both scan-readability and SEO.
Keyword stuffing in headers. Trying to fit the target keyword into every H2 makes the page read robotically. Better to write headers naturally - Google’s semantic understanding catches the topic from natural prose better than from forced keyword placement.
An example
A B2B SaaS team had a 2,800-word pillar article stuck on page 2 for a high-value query. The substance was solid.
The audit found a structural issue: H1 jumping straight to eight H3s in a row, no H2s. The H3s were also generic (“First Point,” “Second Point”).
The rewrite kept substance identical, restructured headers properly (H2-H3 hierarchy, descriptive titles like “How most teams get this wrong” instead of “Common Mistakes”). Six weeks later: ranking moved from page 2 to position 4. CTR lifted because descriptive section titles appeared in some snippet variants. Same content, better structure.
We built Penfriend to handle header-tag structure correctly by default - one H1 per page, logical H2 hierarchy, descriptive rather than keyword-stuffed. Header hygiene is invisible when right and obvious when wrong; we treat it as table stakes.
Related terms
- Featured Snippet - the SERP feature header structure heavily influences candidate selection for
- Anchor Text - the on-link signal that pairs with header structure for SEO
- Canonical URL - the SEO directive that determines which page’s headers Google indexes
- Copywriting - the craft most affected by header structure choices
- Accessibility Terms - the constraint set header tags exist partly to serve
