This factor counts how many times terms from the run’s generated LSI list appear in a page’s H3 headings. It returns a whole-number occurrence count, and it tracks how closely your third-level subheadings echo the related language RankGear derived for the query — it does not, on its own, move a page up the results.
| Factor ID | RG-LSI-004 |
|---|---|
| Family | Related Terms (LSI) |
| Measurement | Related-term occurrence count |
| Measured zone | H3 text |
What it measures
RankGear counts run-specific LSI term occurrences inside the page’s H3 headings. The LSI list is generated per run from the analyzed result set, so the same page can score differently against different queries. A higher count means your H3 subheadings use more of the related vocabulary RankGear associated with the topic.
How RankGear measures it
RankGear parses the text of every H3 on the page and sums case-insensitive, word-aware matches for each item in the generated LSI list. Every occurrence is tallied, including repeats of the same term across different headings. Because at least one term always survives list generation, no list produces a count of 0.
What RankGear looks for
<h3>Choosing the right running shoe</h3>
<h3>Arch support and cushioning</h3>
LSI list: [running shoe, arch support, cushioning, ...]
Match: whole words, case-insensitive, summed across all H3 textHow to optimize it
Name supporting subsections with relevant related language when it genuinely reflects the content beneath each H3. Treat the count as an observation of how well your subheadings mirror the topic’s related vocabulary, not a target to inflate — padding H3s with list terms that do not match the section below degrades the reading experience without describing your content honestly.
Important considerations
- Repeated occurrences all count, so the same term used across several H3s adds to the total.
- The score depends on the LSI list generated for that specific run; a different query or result set can change it.
- Matching is word-aware and case-insensitive, so it keys on whole words rather than substrings or capitalization.
- A strong H3 may use valid, accurate terminology that simply is not present in the generated list, and will not be credited for it.
- This is a correlation-based prioritization signal that relates to position within a measured result set — it does not make a page rank or guarantee any ranking.
Related factors
Part of the Factors reference · how RankGear measures · glossary.