Explain TF-IDF and its use in SEO analysis

WHAT THIS TESTS: It checks that you understand the TF-IDF formula and can wield it as a practical content-gap tool against the pages already ranking for your target query.

A GOOD ANSWER COVERS: Term frequency measures how often a term appears in one document, optionally length-normalized. Inverse document frequency is the log of the corpus size divided by the number of documents containing the term, so words appearing everywhere score near zero. Multiplying them gives a weight that is high only for terms frequent in a document yet uncommon across the corpus, making them distinctive. For SEO topic analysis, you scrape the top-ranking pages for a target query, treat them as your corpus, and compute TF-IDF across them. Terms with consistently high weights are the vocabulary the topic demands. Comparing your draft's terms to that set exposes relevant concepts you have omitted, guiding additions that improve topical comprehensiveness rather than arbitrary repetition.

COMMON WRONG ANSWERS: Calling it a keyword-density score to hit a fixed percentage, which invites stuffing. Confusing the IDF direction so common words look important. Forgetting that the corpus choice, the top-ranking pages, is exactly what makes the analysis competitive rather than absolute against generic text.

LIKELY FOLLOW-UPS: How do you assemble and clean the corpus of competitor pages. Why use a logarithm in IDF rather than a raw ratio. How does this classic technique relate to modern embedding-based relevance models. How do you avoid over-optimizing a page until it reads unnaturally.

ONE CONCRETE EXAMPLE: Analyzing the top ten pages for 'sourdough starter,' TF-IDF surfaces distinctive terms like 'hydration,' 'discard,' and 'levain' that recur across competitors but are rare in general text. If your draft never mentions 'hydration,' that is a concrete, defensible content gap to address, not an arbitrary keyword to repeat a fixed number of times to chase a density figure.