This factor counts how many keyword variations appear inside your HTML table <caption> elements. It is a count of matches within captured caption text, not a ranking guarantee — a page with no data tables can score zero and still rank perfectly well.
| Factor ID | RG-KWD-041 |
|---|---|
| Family | Keyword Usage & Density |
| Measurement | Variation occurrence count |
| Measured zone | Raw captured caption contents |
What it measures
This factor counts keyword variations that appear inside HTML table captions. A table <caption> is the short label a data table carries to describe what it holds, and this factor totals how many recognized variations of your target terms surface in that caption text across the page.
How RankGear measures it
RankGear extracts the recognized caption contents from the script-free body without additional cleaning, then totals whole-word variation matches case-insensitively. Because the capture is taken as-is, any nested markup inside a caption stays in the captured text. The result is a single count of variation occurrences found within table captions.
What RankGear looks for
<table>
<caption>Quarterly organic traffic by channel</caption>
<tr><th>Channel</th><th>Sessions</th></tr>
...
</table>How to optimize it
Give data tables concise, descriptive captions that state what the table represents, and let relevant terms appear naturally when they genuinely fit the label. Treat the count as an observation of how your tables are described, not a target to inflate — stuffing variations into a caption works against the clarity a caption exists to provide, and only tables carry captions in the first place.
Important considerations
- This measures table
<caption>elements specifically, not image captions or figure captions in general. - Nested markup inside a caption remains in the captured text and is not stripped before counting.
- Pages without tables may correctly score zero — the absence of a caption count is not a defect.
- Accessibility and clarity are the primary purposes of a caption; describing the table well matters more than the variation count.
- This is a prioritization signal that relates to position within a measured result set, not a cause of ranking — correlation is not causation.
Related factors
Part of the Factors reference · how RankGear measures · glossary.