Analyze term frequency-inverse document frequency for content optimization. Compare your content terms against corpus benchmarks.
Drop in a draft and a set of comparison documents - competitor posts, category benchmarks, or your own archive - and this tool scores every term with the tf-idf formula: how often a word appears in your text, weighed against how rare it is across the corpus. The analysis runs entirely in the browser; pasted drafts never leave the machine. The result most people misread: topic words like 'marketing' or 'blog' often score near zero because every benchmark document contains them, while oddly specific words rank highest. The output is a term-weighting lens, not a keyword-rank guarantee.
A concrete input and expected output from the current implementation.
Input
draft: email marketing for small business corpus: 1. email marketing campaigns for small business 2. social media marketing for restaurants 3. email marketing for ecommerce startups
Expected output
small 1.099 business 1.099 email 0.405 marketing 0.000 for 0.000
With a 3-document corpus, idf = ln(3 / documents containing the term). 'small' and 'business' occur in one corpus document (ln(3/1) ~ 1.10), 'email' in two (ln(3/2) ~ 0.41), and 'marketing' and 'for' in all three (ln(3/3) = 0), which zeroes their weight even though the draft is about marketing.
Term frequency is how often a word appears in your document; inverse document frequency is ln(total corpus documents / documents containing the word). Their product is the weight: a word in every corpus document scores zero no matter how often you use it, and a word in just one document scores the maximum. Both numbers come from the exact text you pasted, so spelling, casing, and tokenization change the results.
Every score is relative to the documents you supply. Against a broad, general corpus most terms look distinctive; against a narrow corpus of near-identical posts, the same draft looks ordinary. The identical text therefore scores differently on each run, so build the corpus to mirror what you want to be compared with - competitors, top-performing pages, or a category sample. Stopwords are typically dropped, and punctuation and case are usually normalized before counting.
Because idf is ln(N/df): if the term appears in every corpus document, ln(3/3) is zero, and any frequency in your text multiplies to zero. The score means the term is shared vocabulary in your benchmark set, not that it is wrong to use. Against a broader corpus the same term scores above zero.
No. The score measures distinctiveness against the corpus you supplied, nothing about reader demand. Piling on high-scoring off-topic words reads as keyword stuffing and lowers quality. The reliable use is confirming the terms you already chose are the distinctive ones, and spotting a missing topical term from a competitor set.
The example and behavioral notes were checked against the browser implementation. Standards and primary references below define the relevant format, formula, or platform behavior.
Shorten URLs with local redirect tracking via localStorage. See click counts and generate a QR code for each link.
Build UTM-tagged URLs for campaign tracking with source, medium, campaign, term, and content parameters.
Enter 2+ headlines and see simulated previews in Google results, email subject lines, and social posts side-by-side.
Count characters for Twitter/X, LinkedIn, Facebook, and Instagram with platform-specific limits and warnings.
Calculate click-through rates from impressions and clicks with industry benchmarks and trend analysis.
Check email subject line length against best practices for desktop and mobile with preview text optimization.