Exact Matches in Top 30KB counts how many times your exact target query appears, case-insensitively, near the start of a page. RankGear takes the first 30,000 UTF-16 units of the raw source, pulls the image alt text found inside that span, strips selected noncontent elements, appends those alt values, and tallies exact-phrase matches. It is an occurrence count observed within a measured result set, not a quota to hit — more matches do not on their own cause or guarantee a ranking.
| Factor ID | RG-KWD-013 |
|---|---|
| Family | Keyword Usage & Density |
| Measurement | Exact-query occurrence count |
| Measured zone | Truncated Google-density corpus plus image alt text |
What it measures
This factor counts exact-query occurrences inside RankGear’s compatibility view of the beginning of the page source. Rather than reading the whole document, it looks at the opening span that a search engine weighs most heavily and reports how many times the exact query phrase appears there — a raw tally, not a percentage.
How RankGear measures it
RankGear truncates the raw HTML to 30,000 UTF-16 units, extracts the alt text present within that truncated source, strips selected noncontent elements from what remains, appends those alt values, and performs case-insensitive exact-phrase matching across the resulting corpus. The order matters: truncation happens before stripping, and only alt text that falls inside the truncated source is appended.
| Step | What happens |
|---|---|
| 1. Truncate | Raw HTML is cut to the first 30,000 UTF-16 units. |
| 2. Extract alt | Alt text is read from that already-truncated source only. |
| 3. Strip | Selected noncontent elements are removed from the truncated source. |
| 4. Append | The extracted alt values are added to the stripped corpus. |
| 5. Match | Case-insensitive exact-phrase matches of the query are counted. |
How to optimize it
Introduce the topic clearly in useful early-page content, and write accurate alt text for relevant images that appear within the measured source range. Treat this value as an observation of how plainly the page opens on its subject, not a target to engineer toward — padding the opening or stuffing phrases into alt text to move the count degrades the page without earning the position.
Important considerations
- Truncation happens before stripping for this factor, so the measured span is fixed by source length — not by how much visible prose it contains.
- Only alt text present inside the truncated source is appended; images whose markup falls beyond the 30,000-unit cut do not contribute.
- The stripped corpus can still contain markup tokens, which sit alongside the readable text that is matched.
- Page templates and source order can affect the result, since they change what lands inside the truncated span.
- This is a count that relates to position within a measured result set — it does not on its own cause or guarantee a ranking.
Related factors
Part of the Factors reference · how RankGear measures · glossary.