N-gram Keyword Extractor

Single word counts hide the topic

Counting individual words tells you a page mentions "running" forty times and "shoes" thirty five times. It does not tell you whether the page is about running shoes, shoe running, or two unrelated subjects sharing a page. Phrases of two, three and four words carry the meaning that individual words lose, and extracting them shows what a piece of content is actually about rather than what its author believes it is about. The gap between those two is frequently the whole problem.

Stop words have to go, but only from the edges

Removing every occurrence of "of" and "the" destroys genuine phrases such as "rate of return" and "the state of the art". The workable rule is to discard phrases that begin or end with a stop word while keeping those where one sits in the middle. That keeps meaningful multi-word terms intact while removing the enormous number of grammatical fragments that would otherwise dominate any frequency list.

Frequency is a description, not a target

Knowing which phrases repeat is diagnostic. It is not an instruction to repeat them more. Keyword density as a ranking factor has not been meaningful for many years, and writing to hit a density figure produces text that reads badly to the only audience that matters. The useful readings are different: a phrase you expected to appear and do not, a phrase dominating that you did not intend, and the vocabulary a page uses compared with one that already ranks.

Related tools

• Meta Tag Analyzer — Check a page's meta tags for the problems that actually affect indexing and click-through.
• Redirect Chain Checker — Trace the full path of redirects (301/302) to find loops or lost link juice.
• Robots.txt Validator — Validate robots.txt syntax and check whether a specific URL is allowed or blocked for a given crawler.
• Sitemap Generator (Lite) — Generate a valid XML sitemap with lastmod dates and correct structure, and learn which URLs belong in it and which do not.