Back to Tools

Keyword Analyser

On-Page SEO, Content & Term Association
Score — 0 words 0 terms

Source & Settings

URL Mode
Compare with competitor pages (optional)

Each page is fetched and analysed, then compared with yours to find topics they cover that you don’t.

Analysis Options

Analysis & Insights

Waiting for input

Ready to analyse

Enter a public URL in the left panel, or paste an article or its HTML. Add a target keyword to see where it appears on the page.

About Keyword Analyser

Documentation

Keyword Analyser is a comprehensive on-page SEO and lexical analysis studio engineered to evaluate webpage clarity, verify on-page search ranking signals, map term co-occurrence networks and uncover content opportunities against competitor pages.

Overview at a glance

Keyword Analyser platform specifications
Application type Browser-based SEO auditing and text analysis studio
Core processing 100% client-side natural language processing engine (tokenisation, stemming, readability, term association)
CORS proxy Netlify serverless edge proxy with comprehensive headers, timing and metadata inspection
Supported inputs Public webpage URLs, raw pasted HTML markup and plain text copy
Readability tests Flesch Reading Ease score, US school grade level, sentence length distribution and syllable counts
Lexical diversity MATTR (Moving-Average Type–Token Ratio) across 50-word sliding windows
Term weighting Zipf-bucketed TF-IDF frequency calculation derived from Speer wordfreq corpus (19,846 words)
Graph algorithm Interactive HTML5 canvas force-directed graph with modularity-based topic clustering
Search intent Automated 4-way classification (Informational, Commercial, Transactional, Navigational)
Competitor benchmarking Multi-URL simultaneous comparison extracting content gaps and unique topical strengths
Licence MIT

What Keyword Analyser does

Keyword Analyser bridges the divide between technical search engine audits and editorial content analysis. Rather than simply counting raw word occurrences, it evaluates the structural, linguistic and contextual health of web content.

Technical on-page verification

The engine audits core HTML signals including page title length, meta description snippet boundaries, H1 and subheading hierarchy, canonical URL integrity, mobile viewport tags, Open Graph previews and Schema.org structured data validity.

Readability and prose metrics

It computes the standard Flesch Reading Ease score alongside US educational grade levels, tracking sentence lengths and polysyllabic word ratios to ensure content aligns with the reading habits of your target audience.

Lexical variety and repetition monitoring

By applying a Moving-Average Type–Token Ratio (MATTR) over sliding 50-word windows, the analyser assesses vocabulary richness independently of document length and detects excessive keyword repetition before it risks reading as spam.

Interactive co-occurrence networks

The built-in canvas network visualises how terms interconnect within individual sentences, clustering related concepts into topical communities that demonstrate topical authority.

Competitor content gap analysis

By benchmarking up to four competitor URLs alongside your primary page, the studio extracts topics and named entities your rivals cover that are missing from your article.

How Keyword Analyser works

  1. Input page address or draft copy. Provide a live URL or paste raw HTML or plain text into the studio input panel.
  2. Retrieve and extract content. Webpages are fetched through the Netlify edge proxy, isolating primary article prose from headers, sidebars, navigation menus and footers.
  3. Lexical tokenisation and filtering. Prose is split into clean tokens. Stop words are optionally stripped, and the Porter stemming heuristic unifies morphological forms.
  4. Metric calculation. The engine computes readability formulas, MATTR diversity scores, Zipf-based TF-IDF weights and search intent classifications.
  5. Network synthesis. Sentence co-occurrences are mapped into a force-directed layout graph with Louvain modularity clustering.
  6. Prioritised guidance. The studio synthesises findings into an actionable list of fixes ordered by organic search impact.

Who Keyword Analyser is for

Technical SEO consultants

Auditors needing instant verification of on-page tags, canonical headers, HTTP response timings and Schema structured data without heavy desktop crawlers.

Content strategists and copywriters

Writers who want to ensure their prose is easy to read, uses rich vocabulary and covers all related terms naturally without awkward keyword stuffing.

Digital marketers and growth leads

Marketers looking to identify content gaps between their high-priority landing pages and the leading organic search competitors.

Web developers

Engineers checking that dynamic page templates produce valid heading outlines, clean server responses and correct social meta tags.

SEO and content metrics explained

Flesch Reading Ease

Scored on a scale of 0 to 100. Higher numbers represent easier reading. Scores of 60 to 70 represent standard plain English accessible to 13 to 15 year olds, whilst scores below 50 reflect academic or technical text.

Moving-Average Type–Token Ratio (MATTR)

A sliding 50-word window calculates the proportion of distinct words, generating an average ratio across the entire text. Published prose typically scores 80% or higher.

TF-IDF and Zipf background frequency

Terms are evaluated against 19,846 common English words from the Speer wordfreq corpus. Words that appear frequently in your document but rarely in general English receive the highest TF-IDF weight.

Search intent classification

Signals across headings, button labels and schema markup classify the dominant query intent into Informational (guides, tutorials), Commercial (reviews, comparisons), Transactional (prices, checkouts) or Navigational (brand pages, account portals).

Technical specifications

Application type Browser-based SEO auditing and text analysis studio
Core processing 100% client-side natural language processing engine (tokenisation, stemming, readability, term association)
CORS proxy Netlify serverless edge proxy with comprehensive headers, timing and metadata inspection
Supported inputs Public webpage URLs, raw pasted HTML markup and plain text copy
Readability tests Flesch Reading Ease score, US school grade level, sentence length distribution and syllable counts
Lexical diversity MATTR (Moving-Average Type–Token Ratio) across 50-word sliding windows
Term weighting Zipf-bucketed TF-IDF frequency calculation derived from Speer wordfreq corpus (19,846 words)
Graph algorithm Interactive HTML5 canvas force-directed graph with modularity-based topic clustering
Search intent Automated 4-way classification (Informational, Commercial, Transactional, Navigational)
Competitor benchmarking Multi-URL simultaneous comparison extracting content gaps and unique topical strengths
Licence MIT

Privacy, security and architecture

Keyword Analyser was built with a strict privacy-first architecture. All text analysis, tokenisation, readability calculations and graph visualisations run entirely within your local browser.

When fetching remote URLs, the request is routed through a Netlify edge function solely to overcome browser cross-origin resource sharing (CORS) constraints. The proxy does not log, track or retain the content of requested webpages.

Glossary of terms

TF-IDF
Term Frequency–Inverse Document Frequency. A statistical measure evaluating how important a word is to a document relative to an authoritative corpus of common language usage.
Zipf Frequency
A logarithmic metric measuring word frequency per billion words. High values denote ubiquitous function words, whilst lower values identify distinctive topical terms.
Flesch Reading Ease
A standardised readability formula scoring text between 0 and 100 based on average sentence length and syllable count per word. Higher scores indicate greater clarity.
MATTR
Moving-Average Type–Token Ratio. A robust vocabulary diversity metric that calculates unique word ratios across fixed sliding windows, avoiding document length distortion.
Stemming
An algorithmic heuristic (such as the Porter stemmer) that strips morphological suffixes to consolidate singular, plural and grammatical variants into a root form.
Stop Words
High-frequency function words such as articles, conjunctions and prepositions that carry minimal standalone topical meaning in keyword frequency calculations.
Co-occurrence Graph
A network diagram connecting words that frequently appear together within identical sentences, revealing underlying topical clusters and entity relationships.
Search Intent
The primary underlying motivation of a search engine user, classified into informational, commercial, transactional or navigational categories.
Canonical URL
An HTML link element specifying the authoritative, preferred URL of a webpage to prevent duplicate content indexing issues across search engines.
Meta Description
An HTML header element providing a concise summary of a webpage, frequently utilised by search engines to populate organic search snippet previews.
TTFB
Time to First Byte. The duration in milliseconds from the initial HTTP request until the client browser receives the very first byte of response data from the host server.
JSON-LD
JavaScript Object Notation for Linked Data. A machine-readable format embedding Schema.org structured data to enable rich snippets and entity understanding.

Frequently asked questions

What is Keyword Analyser?

Keyword Analyser is a comprehensive on-page SEO and text analysis studio. It evaluates webpage readability, inspects technical SEO signals, identifies term associations through interactive force-directed graphs and highlights content gaps against competitors.

How does Keyword Analyser fetch live webpages?

When you enter a public URL, the studio requests the page via a dedicated Netlify serverless edge proxy. The proxy retrieves the raw HTML along with complete HTTP headers, redirects and response timing, returning it to your browser for client-side evaluation.

Can I analyse draft copy or pasted HTML markup without a public URL?

Yes. Switch to the Paste text or HTML tab to inspect unreleased drafts, blog posts or raw source code. The analyser executes full content, readability and keyword calculations immediately without any network requests.

What does the SEO Health Score measure?

The SEO Health Score is a weighted composite index between 0 and 100. It measures content depth and readability, technical on-page health (title, meta description, heading structure, canonicals, mobile viewport) and target keyword alignment where specified.

How is the keyword co-occurrence graph constructed?

The engine examines the proximity of words across individual sentences. Words appearing together in multiple sentences receive connecting edges weighted by mutual occurrence. A Louvain community detection algorithm then identifies and colour-codes topical clusters.

What is the advantage of MATTR over standard Type-Token Ratios?

Standard Type-Token Ratios decrease naturally as text gets longer because natural language repeats core words. Moving-Average Type–Token Ratio (MATTR) samples a sliding 50-word window throughout the document, providing an unbiased assessment of vocabulary diversity.

How does competitor gap analysis work?

You can supply up to four competitor URLs alongside your primary page. The studio fetches each page, extracts its dominant keywords and named entities, and isolates topics and entities covered by your rivals that are absent from your page.

Are my analysed URLs or pasted texts stored on any server?

No. All parsing, lexical calculations and keyword extractions execute entirely in your browser using JavaScript. The fetch proxy forwards webpage HTML without saving, logging or persisting any content.

Can I export the analysis findings?

Yes. You can export complete keyword frequency tables as CSV or JSON files, copy formatted Markdown summary reports to your clipboard, generate shareable links for public URLs or print a formatted PDF audit.

Does Keyword Analyser support stemming?

Yes. Enabling Group word forms consolidates grammatical variants, plurals and conjugated verbs into their root lexical form, ensuring that variants such as optimisation, optimising and optimise count toward a single topical score.

About the author

Keyword Analyser was engineered by Mike Higgins, a technical SEO and web operations manager with over a decade of experience accelerating digital growth and building performance-driven web tools.

Explore more developer and SEO tools on the Tools index or visit the About page to learn more.