Paste HTML to list matching H1–H6 opening tags, count H1 matches and flag downward level jumps. Review the findings as a pattern-based check, not a full page audit.
The heading structure checker scans pasted text for opening tags that resemble H1 through H6. It lists those matches in source order, counts H1 matches and flags certain jumps between adjacent levels. Paste the relevant HTML and select Check Headings to create a report. It does not fetch a live URL, render the document or decide which heading a visitor actually sees after styles and scripts are applied.
Choose a sample that matches the question you want to answer. A full page includes shared navigation, sidebars and footer regions, while a content fragment may intentionally omit the page's main heading. If the fragment report says H1 is missing, that does not prove the complete page lacks one. Keep the scope of the input with the result so a reviewer can distinguish a content-snippet check from a claim about the entire rendered document.
The function searches case-insensitively for a less-than sign, h, a digit from one to six and characters through a closing angle bracket. It does not build a browser document tree or validate that a matched tag has an appropriate closing partner. An opening heading tag with attributes is listed as written. The report shows the matched tag syntax, not the heading's visible text content, so you still need the source to assess its wording.
Because this is a text pattern, heading-like examples inside comments or script strings can also match. A malformed tag whose name begins with a supported pattern can be counted in ways a real HTML parser would not treat as a heading. Treat the output as a quick search through supplied markup, not proof of valid HTML. For a production audit, compare the findings with the browser's actual document structure and the section the heading is intended to label.
A sequence of H1, H2 and H3 produces no downward-jump warning because each adjacent descent increases by one level. H1 followed directly by H3 produces a warning for the skipped H2 level. The check compares each matched tag with the previous one in the source stream; it does not recognize independent section trees, template boundaries or a semantically separate region that might explain the source ordering.
Moving back from H3 to H2 is not flagged by this rule, because the numeric level is decreasing rather than jumping downward into a deeper level. Likewise, H2 followed by H4 triggers a warning even if another part of the page uses H3 elsewhere. These examples show why a warning is a prompt to inspect a local sequence. It is not a complete interpretation of every relationship between sections or a guarantee that an unflagged outline is understandable.
The interface describes zero H1 matches as missing, more than one as multiple and exactly one as good. Those are simple status labels based on a count. They do not verify whether the H1 accurately describes the page, whether it is hidden, whether the markup is valid or whether the rest of the content meets the reader's needs. Do not read “Good SEO” in the report as a verified ranking assessment or a complete accessibility verdict.
For your own content, a clear main heading and meaningful section organization are useful editorial goals. Review multiple H1 findings in context instead of assuming the number alone proves a penalty. A repeated header component, a pasted example or an incomplete snippet can explain the count. Resolve the actual structural or wording issue rather than changing tags only until the status turns green without understanding what the final document communicates to users.
A perfectly sequential set of heading levels can still have unhelpful labels such as “More”, “Section” and “Details” repeated throughout a page. The checker does not inspect the text between the tags, compare headings with their following paragraphs or identify whether a label describes its section accurately. Read the actual headings as a short outline and ask whether a new reader could predict what each part of the document covers.
Use levels to express a relationship between the main subject, its sections and their subsections. Choose descriptive wording for the content rather than selecting a level solely because its default font size looks right. Styling and document structure can be handled separately in the site design. This review goes beyond what the pattern report calculates, so a change in heading wording may improve navigation even if the reported tag sequence remains exactly the same afterward.
A JavaScript application may create or change headings after the initial HTML loads, and CSS can hide a matched element or visually reorder sections. Pasting only the original response does not capture every later change. Conversely, copying rendered text without tags supplies nothing for this matcher to recognize. Decide whether you are reviewing server output, a template fragment or the current DOM, then collect the corresponding HTML for the check.
For a practical browser review, inspect the relevant rendered elements and compare them with the source report. Check whether the heading is visible, labels the intended region and appears in a sensible reading order. The checker does not evaluate keyboard navigation, accessible names or how assistive technology presents the page. Use appropriate browser and accessibility inspection tools when those questions matter instead of extending a narrow pattern result into an unsupported full-page assessment.
If the same warning appears on many pages, inspect whether a shared header, card or sidebar introduces the repeated tag pattern. Changing one shared template can affect many routes, so review several representative page types before applying a blanket level replacement. An article page, a tool page and a category page can have different structures even when they reuse components. A consistent fix should reflect those actual relationships rather than a directory-wide text replacement.
Keep a short before-and-after sample with any proposed change. Note which sequence triggered the warning, what section relationship you intended and what the rendered page shows afterward. This gives a reviewer enough evidence to assess the change beyond a color or H1 total. Do not duplicate a main heading in the explanatory content merely to satisfy a local fragment check if the full page already supplies the appropriate label through its template.
Select Check Headings again after changing the input, because the report represents the last explicit run. Compare the found tags with the edited source and investigate any unexpected match in comments, examples or malformed markup. Empty input and input without heading-like tags produce different messages; neither result can establish what a separate live page contains unless that page's relevant markup was actually supplied to the tool.
Save the checked sample, report and page context in your normal workflow if the result supports a content or template review. Then inspect the deployed output after integration rather than assuming the draft snippet was used unchanged. The useful outcome is a readable structure supported by the actual HTML, not simply a report with no warnings. Treat this page as one focused inspection aid alongside content review, valid markup and a real rendering check.
It matches H1–H6-like opening tags in pasted text, lists them in source order, counts H1 matches and reports certain level jumps.
No. The tool expects pasted HTML and does not fetch or render a website.
No. It lists matched opening tags. Review the source or rendered page to inspect the actual labels.
A warning appears when an adjacent matched level is more than one step deeper, such as H1 followed by H3.
No. The rule flags deeper jumps, not moves back to a shallower level.
No. That interface message reflects a count, not a verified ranking, content-quality or accessibility assessment.
Yes. The search is pattern-based and does not exclude every context a proper document parser would distinguish.
No. Matching opening tags does not establish that they are paired correctly or that the complete HTML is valid.
The full page may provide its main heading elsewhere. Interpret the report according to the exact snippet supplied.
No. Select Check Headings again after an edit to inspect the new input.