Contextual Density (Unique Terms) is a unique-term rate: it counts how many distinct query-specific variations, entities, and related terms appear on the page at all, divides that by the page’s meaningful-word count, and multiplies by 100. Each term counts once no matter how often it repeats, so this reads topical breadth rather than emphasis — and it is a comparative rate that can exceed 100, not a bounded percentage.
| Factor ID | RG-KWD-100 |
|---|---|
| Family | Keyword usage & density |
| Measurement | Unique-term rate |
| Measured zone | Cleaned rendered page content |
What it measures
This factor measures the breadth of query-specific variations, entities, and related terms present on the page relative to its meaningful-word count. Each distinct source-list entry contributes at most once within that list, regardless of how many times it appears on the page. The intent is to separate topical breadth — how many different relevant concepts the page touches — from the repeated-occurrence emphasis captured by occurrence-based Contextual Density.
How RankGear measures it
RankGear analyzes the cleaned page content and checks whether each distinct variation, entity, and LSI or related term is present. Presence is determined with lowercased substring matching, and each present entry is counted once within its source list. RankGear then subtracts English stop words from the page’s total word count, divides the combined number of present terms by the remaining meaningful-word count, and multiplies by 100. If no meaningful-word denominator remains, the value is 0.
rate = ( present terms across all source lists
÷ (total words − English stop words) )
× 100The same wording can contribute once in more than one source list, and because the check is substring-based it can match a shorter string inside a longer word. The result is therefore a comparative rate rather than a bounded percentage, and it can exceed 100 in unusual cases.
How to optimize it
Treat the value as an observation about breadth, not a target to inflate. Use the run’s topical lists to find important concepts that are absent from the target page, then expand the content with distinct, relevant ideas that improve the answer: named entities, terminology, attributes, processes, use cases, alternatives, and the relationships users need to understand.
Do not paste terms into the page as a glossary without context. Integrate them into accurate sentences, headings, examples, tables, or explanations that make the page genuinely more complete — the goal is a better answer, not a higher count.
Important considerations
- This factor measures presence breadth, not semantic correctness or factual quality.
- Substring matching can create occasional apparent matches inside longer words.
- Provider availability affects the entity list and can change the measurement.
- Compare pages only within the same run and configuration.
- Review it alongside occurrence-based Contextual Density to see whether a page is broad, repetitive, both, or neither.
- The rate relates to position within a measured result set; it is a prioritization signal, not proof that added terms cause a page to rank.
Related factors
Part of the Factors reference · how RankGear measures · glossary.