TF-IDF Calculator: Find the Terms Your Content Is Missing

TF-IDF, short for term frequency-inverse document frequency, is one of the oldest and most reliable ways to measure which words actually matter to a piece of text within a set of comparable documents. Modern content optimization tools use a version of this same math to compare an article against the pages already ranking well for a topic, surfacing terms those pages use heavily that a draft might be missing. This calculator runs the same idea entirely in your browser: paste your content and one or more competitor texts, and it returns the terms competitors emphasize the most relative to you.

How TF-IDF is actually calculated

Term frequency (TF) is simple: how often a word appears in a document divided by that document's total word count. A word used 8 times in a 400-word article has a TF of 0.02. Inverse document frequency (IDF) adjusts for how common that word is across every document in the set, calculated as the logarithm of the total number of documents divided by how many of them contain the word. A term that shows up in every document gets an IDF close to zero, because it is not distinctive to any one of them. A term that appears in only one or two documents out of several gets a higher IDF, because its presence is more specific. Multiplying TF by IDF gives the final score: high when a term is used heavily in one document and rarely elsewhere in the set, low when it is either barely used or used everywhere equally.

Reading the gap column

This tool calculates a TF-IDF score for your content and for each competitor text you paste, then averages the competitor scores per term and sorts by the difference. A large positive gap means competitors are, on average, giving a term much more weight than your content does. That is not proof your content is worse, but it is a strong prompt to check: does your article actually cover that subtopic, or has it been skipped entirely? A term like "interval workouts" scoring high for competitors but near zero for you, in a piece about beginner running plans, is worth investigating even if you never repeat that exact phrase.

Using the results without stuffing

The mistake to avoid is mechanically inserting every high-gap term into your text without adding real information. TF-IDF measures emphasis, not magic words, and a term repeated without substance behind it will not reproduce whatever depth made competitors' scores high in the first place. The better approach is to treat each high-gap term as a question: is there a subtopic, a use case, or a detail here that competitors cover and you do not? Add that coverage in your own words, and let the natural repetition of genuinely relevant terms follow from actually writing about the thing. For a more reliable read, paste in two or three real competitor pages rather than one, since a single comparison document can make almost anything you do not share look artificially important. If you want content built around real topical coverage rather than a single metric, that is exactly what our content optimization team does for a living.