RG-KWD-098 counts the u (underline) elements on a page whose entire cleaned text is exactly one of the run’s keyword variations — nothing more, nothing less. Any extra word beyond the variation breaks the match, and a higher count relates to position within a measured result set rather than causing a page to rank.
| Factor ID | RG-KWD-098 |
|---|---|
| Family | Keyword Usage & Density |
| Measurement | Whole-element variation match count |
| Measured zone | Cleaned u contents |
What it measures
This factor counts the u elements whose entire cleaned text equals one keyword variation. It is a whole-element test: the underlined text must be a variation in full, not merely contain one. A u element that carries the variation plus any additional wording is not counted.
How RankGear measures it
RankGear extracts and cleans the bounded contents of each u element, then tests each one against a case-insensitive variation expression that is anchored at the start and the end. Because the expression is anchored on both sides, only a u element whose complete cleaned text is a variation — with no leading or trailing text — produces a match.
What RankGear looks for
<u>trail running shoes</u>
matched against /^(best running shoes|running shoes for men|trail running shoes)$/iIf the u content is… | What RankGear records |
|---|---|
| Exactly a variation, e.g. Trail Running Shoes | Counts toward the factor |
| A variation plus extra words, e.g. the best trail running shoes | Does not count — additional text prevents the match |
Underlined only by CSS, with no u element | Does not count — visual underlining is not measured |
How to optimize it
Use a variation-only u element only when underlining or annotation is genuinely appropriate for the reader and will not be mistaken for a link. Treat this as an observation about how your underlined annotations align with a target variation, not a quota to hit — never underline a phrase for the sole purpose of moving this value. If underlining does not serve the reader, leave it out.
Important considerations
- Additional text prevents a match: even a single extra word inside the
uelement breaks it, because the pattern is anchored at both ends. - CSS underlining does not count — only a real
uelement with matching cleaned text is measured, not text styled to look underlined. - Variation alternatives are not regex-escaped, so variations containing regex metacharacters may behave unexpectedly.
- Do not underline phrases solely to influence this value; underlining that misleads readers into expecting a link works against the page.
- This is a correlational prioritization signal. A higher count tracks with position inside a measured result set; it does not make a page rank or guarantee any position.
Related factors
Part of the Factors reference · how RankGear measures · glossary.