Heading tags (h1 through h6) exist to create a logical outline of a page's content, similar to a table of contents. When that outline is broken — no h1 at all, several competing h1s, or a jump straight from h2 to h5 — it's harder for both a screen reader user navigating by headings and a search engine trying to understand the page's structure.
One h1, describing the whole page
A page should generally have exactly one h1, summarizing what the page is about — the way a book has one title, not several. Multiple h1 tags don't reliably cause a penalty, but they do dilute the single clearest signal a page could otherwise send about its main topic.
Skipping levels breaks the outline, even visually
Jumping from an h2 straight to an h4 without an h3 in between might look fine visually if the styling happens to match, but it breaks the actual document outline that screen readers and some SEO tools rely on — the equivalent of an outline going "1, 2, 2.1.1" with no "2.1."
TeckForge's Heading Structure Checker extracts a page's full heading outline from its HTML, flags missing or duplicate h1 tags and skipped levels, and shows the resulting structure as a readable, indented outline.