How this tool works
The analyzer tokenizes the text you paste, counts words and characters, estimates sentence and paragraph totals, and compares up to 25 terms using exact phrase or individual-word matching. Keyword share is a descriptive ratio of matches to total word tokens; it is not a target or ranking score.
How to use it
- Paste the text you want to review.
- Optionally enter up to 25 words or phrases separated by commas.
- Choose phrase or individual-word matching and set a reading speed if you want an estimate.
- Review the counts and term table, then export a CSV if useful.
- Use the numbers as editing context and keep the wording natural and helpful.
Limits to keep in mind
It does not fetch a URL, inspect the rendered page, judge topical relevance, detect search intent, or predict rankings. Sentence and paragraph counts are estimates, especially for abbreviations, headings, lists, and languages with different punctuation or tokenization.
Frequently asked questions
Is there an ideal keyword density?
No universal target is provided here. The percentage describes the supplied text and cannot determine whether a page is useful or likely to rank.
Can the analyzer check a live URL?
No. It only processes text pasted into the page; it does not request or store a URL or send the text to a server.
What is the difference between phrase and individual-word mode?
Phrase mode counts consecutive matching tokens in the same order. Individual-word mode counts each distinct word in the entered term, without requiring them to appear together.
Why can the sentence count be approximate?
Abbreviations, decimals, headings, and punctuation conventions can make sentence boundaries ambiguous. Treat the count as a quick estimate.
A practical way to review keywords and a draft together
A keyword count is most useful when it answers a small editorial question: does this draft mention the subject clearly, and is a phrase repeated more often than the writer intended? The Keyword & Content Analyzer brings that check together with word, character, paragraph, sentence, and reading-time estimates. Paste the text you want to assess, then decide whether you need a frequency comparison. If you only want basic counts, leave the terms field empty. The analysis takes place in your browser, so the text is not sent to this site for processing.
The report describes the text you provide. It does not inspect a page on your website, read search results, compare competitors, measure search volume, or estimate the chance that a page will rank. Treat it as a quick editing aid. The decision about what belongs in a page still depends on the reader, the topic, the evidence available, and the actual purpose of the content.
What the keyword share means
For a selected term, the analyzer counts matching appearances and divides that count by the number of word tokens detected in the pasted text. It shows that result as a percentage. For example, if a phrase is found four times in a draft containing 800 word tokens, its displayed share is 0.50 percent. This is a description of that particular text and the tool’s matching rules. It is not a recommended range, a score, or a threshold that a page must reach.
A share can change when the draft changes, when a heading is added, or when a different matching mode is selected. It can also be misleading when considered without context: the same number could describe useful terminology in one document and awkward repetition in another. Read a few of the matches in their sentences. Ask whether each use clarifies the subject for a person. If a phrase sounds forced, repeated, or inserted only to influence a metric, rewrite it for clarity rather than trying to preserve the percentage.
Choose the matching method that fits your question
Exact phrase mode looks for the entered words next to one another and in the same order, after the text has been split into word tokens. If the term is “content audit,” a match for “content audit checklist” begins with the same two-word phrase, while “audit of the content” does not. The checker counts overlapping phrase positions too. This can matter for repeated patterns such as “very very very,” where adjacent matches share a word.
Individual-word mode checks the distinct words in each entered item separately and adds their occurrences. Entering “content audit” in that mode reports the combined appearances of “content” and “audit”; it does not require them to appear as a phrase. This is useful for a rough vocabulary check, but it can obscure how the words are used. If the distinction matters, add each word as its own term or switch back to exact phrase mode.
Enter terms separated by commas, with no more than 25 items. Commas therefore act as separators and cannot be used inside a single keyword phrase. Duplicate entries are collapsed according to the capitalization setting. Match capitalization exactly only when differences in case are part of the question, such as checking a product name. For most editorial reviews, case-insensitive matching is easier to interpret.
Read the draft statistics as estimates
Word count uses sequences of letters and numbers, allowing an apostrophe inside a word. Character totals count Unicode code points, so accented letters and many emoji are counted as one visible character. The “without spaces” value removes whitespace before counting. These conventions are stated so that you can compare drafts consistently, but other editors may tokenize text differently. Do not expect exact agreement with a word processor, publishing system, or another online counter.
Sentence totals are approximate. A period can end a sentence, but it can also appear in an abbreviation, decimal, domain name, or initial. Question marks and exclamation marks can also occur in headings or informal copy. Paragraphs are estimated from blank lines; pasted HTML, bullet lists, and text copied from a formatted editor may not preserve paragraph boundaries. A line count is included as another useful view of the pasted text, not as a measure of structure on the finished page.
Average sentence length is the detected word count divided by the estimated sentence count. The longest-sentence figure uses the same boundary estimate. These values can help a writer notice a section that may need a closer read, but they cannot decide whether a sentence is clear. A short sentence can be confusing, and a longer one can be precise. Read the passage aloud or ask a reader to review it when comprehension matters.
Use reading time as a planning aid
Estimated reading time divides detected words by the reading speed you enter. The default is 200 words per minute, and you can adjust it for your audience or use case. The estimate rounds up to a whole minute and shows less than one minute for shorter drafts. It does not account for diagrams, tables, code, unfamiliar terminology, screen readers, pauses, or a reader who is scanning rather than reading every word.
Use the estimate to compare similar drafts or set expectations in an editorial workflow. Do not present it as a promise about how long every visitor will need. If reading time is shown to your own audience, identify it as an estimate and choose a speed that makes sense for the material. Technical instructions, legal information, and material with many examples often take longer than a plain-text word-rate calculation suggests.
A careful editing workflow
Start with the reader’s task. Before entering any keyword, write down what a visitor should understand or be able to do after reading the page. Paste the draft and use the basic totals to catch accidental omissions, unexpected length changes, or unusually dense sections. These numbers are prompts to inspect the text, not goals to hit.
Next, add the main phrase and a small number of meaningful alternatives. Use exact phrase mode to review wording and individual-word mode when you want a broader count of terms. Sort is not required: the report lists the terms with the most occurrences first so that the largest counts are easy to notice. Open the draft and inspect those phrases in context. A term used in a title, explanation, example, and summary may be appropriate when each use serves a distinct purpose. The same term repeated several times in one paragraph may make the passage harder to read.
Make revisions for accuracy and flow. Use ordinary language, explain specialist terms, and keep details that help the reader. Then run the analysis again if the comparison is useful. You can download a CSV record of the term counts, but it contains only the terms and calculated shares; it does not include the text you pasted. Keep the export only if it helps your review process and avoid treating it as evidence of search performance.
Examples of responsible interpretation
Imagine a product guide where the product name appears in the title, the setup instructions, and the warranty section. A relatively high count may be natural because the name is necessary to identify the item. Removing every repetition could make the guide less clear. The useful question is whether each mention helps a person follow the information, not whether a number falls below a universal limit.
Now consider a draft that uses “best freelance platform” in every other sentence even when the paragraph discusses payment settings or profile photos. The checker can show that the phrase is frequent, but it cannot determine why. A human editor can decide whether the phrase sounds unnatural, whether those sections need a more precise label, or whether the phrase should appear only where it accurately describes the topic.
A short landing page may show a larger percentage for two phrase appearances than a long article, simply because its denominator is smaller. That does not prove that the short page needs more words. Give each page the detail its task requires.
Privacy, accessibility, and known limits
This analyzer runs in the browser and does not fetch a web address. Paste only material you are allowed to process in your browser. The text is not submitted to an analysis API by this tool. If you use a shared device, remember that your browser, clipboard, downloads, or device-management software may keep information independently of this page.
Results appear in summary cards and a table. The CSV export uses the same matching rules. If a result looks surprising, check spelling, punctuation, capitalization, and mode. For comparisons, use the same settings each time.
The analyzer does not decide whether an article is complete, original, accurate, trustworthy, or helpful. It cannot assess experience, evidence, source quality, audience expectations, accessibility in the rendered page, or the quality of a business claim. It does not suggest a target density, and it does not generate keywords. Those decisions require context and human judgment. Use the measurements to locate a question worth checking, then review the actual passage and the needs of its reader.
More SEO Tools · Browse the tools catalogue