What Is Heading Hierarchy?
Heading hierarchy is the ordered nesting of H1 through H6 elements that tells a machine how the sections of a page relate to one another.
Heading hierarchy is the ordered nesting of H1 through H6 elements that tells a machine how the sections of a page relate to one another. It is not a set of font sizes. A heading tag carries structural meaning that assistive technology and parsers act on, so picking H3 because it looks the right size breaks something no sighted visitor will ever report.
Why heading order is worth getting right
Headings are the only structural map most pages publish. A screen reader user pulls up a list of them and navigates by that list, the same way you would skim a contents page. If the list is a jumble of levels, that navigation stops working and the user is left scrolling linearly through a page you designed to be skimmed.
The machine-reading case runs alongside it. Systems that extract a section of your page to answer a question need to know where that section starts and ends, and headings are the cleanest boundary available. Bold text sized to look like a heading gives them nothing. That is a content planning problem more than a development one, which is why it is cheaper to settle the structure before anyone writes than to retrofit it across 200 published posts.
How headings are actually interpreted
- The outline comes from tag order, not appearance. Parsers read the sequence of H1 to H6 elements in the source. CSS can make an H4 enormous and an H2 tiny without changing a single thing about what the markup means.
- Assistive technology exposes headings as a navigable list. Screen reader users move between headings with a keystroke and can jump by level. This is the behaviour that skipped levels degrade.
- Nesting communicates containment. An H3 following an H2 is read as a subsection of it. An H3 following an H1 leaves the relationship undefined, and the reader has to guess whether a level was omitted or the content is unrelated.
- Search engines treat headings as descriptive text, not as a ranking control. They help a crawler understand what a section covers. They are one input among many, and no arrangement of them overrides what the body copy says. The strongest of them still sits at the top, which is why the role of the top-level heading gets discussed separately.
- WCAG puts this under a Level A criterion. Success Criterion 1.3.1, Info and Relationships, requires that structure conveyed visually can also be determined programmatically. Headings styled with CSS alone fail it.
Most real-world faults come from templates rather than writers. Widgets, related-post modules and cookie banners inject headings into the outline that nobody editorial ever chose.
| What the markup does | What a parser concludes |
|---|---|
| H1, then H2, then H3 under it | A clean containment tree; sections have known parents |
| H1, then H3, with no H2 | A level is missing; the relationship is ambiguous |
| Sidebar widget titles marked as H2 | Your sidebar is a major section of the article |
| A pull quote marked as a heading | A section exists that has no content in it |
| Bold paragraph text used as a heading | No section boundary exists at all |
The blunt version
Multiple H1s do not hurt your rankings. The HTML specification has never restricted a document to one, Google has said publicly and more than once that using several is fine, and no audit has ever produced evidence of a ranking loss caused by a second H1. It stays on checklists because it is trivial to detect. A crawler counts H1 elements in one line of code, so it becomes an error row, and error rows justify invoices.
Skipped levels are the fault worth fixing, and rankings are not the reason. Jumping H2 to H4 costs a screen reader user their ability to navigate the page by structure. That is a real person, on your site, unable to find a section, and it is a Level A accessibility failure rather than an SEO one. The honest framing is that heading hierarchy is an accessibility job with a modest parsing benefit attached, not an SEO job.
Apply the same test to everything on a technical checklist: who is harmed if this is wrong, and how? For headings the answer is specific and it is not Googlebot. Most of the other items on a standard on-page checklist survive that test on their own merits. The H1 count does not.
Example
Say a software company builds a pricing page in a page builder. The plan names are set as H4 because H4 matched the design, the feature groups underneath are H2 because the designer wanted them larger, and the page title is an H1. The visual result is correct and nobody notices anything wrong. To a screen reader the outline reads as a title, then five second-level sections, then a set of fourth-level items that appear to belong under a third level that does not exist. A user navigating by heading cannot tell which features attach to which plan. Nothing in the SEO reporting changes. The page is simply harder to use for the people who need the structure most.
FAQ
Can a page have more than one H1?
Yes. HTML permits it and Google has stated that multiple H1 elements do not cause a ranking problem. One H1 remains the clearer choice for most pages, because a single top-level heading gives every other section an obvious parent. Treat it as an editorial preference rather than a rule you are breaking.
Does skipping from H2 to H4 hurt rankings?
There is no evidence that it does. The damage lands on people using assistive technology, who navigate by heading level and lose the structure when a level disappears. WCAG Success Criterion 1.3.1 covers this at Level A, the baseline conformance tier. Fix it for that reason, and stop expecting a traffic change afterwards.
Should headings contain keywords?
They should describe the section accurately, which usually puts the relevant words in them anyway. Writing headings backwards from a keyword list produces labels that do not match the content beneath them, and that costs you readers who skim. Accurate beats optimised here, and accurate is generally optimised as a side effect.
Related terms
- Passage Ranking — what a clean heading structure makes possible when Google surfaces one section of a page.
- Organization Schema — structure for machines expressed in markup rather than in the visible document.
- H1 Tag — the single element that carries the most weight in the hierarchy, and the most myth.
If a technical audit lists multiple H1s as an error and says nothing about skipped levels, it was run by a tool and read by nobody. The second fault affects real users; the first affects a spreadsheet.