GEO Score

"GEO" stands for Generative Engine Optimization — making your pages easy for AI assistants (ChatGPT, Perplexity, Google's AI answers) to understand and quote. Where classic SEO earns you a spot in a list of links, GEO earns you a place inside the answer the AI writes.

The GEO score measures how quotable a page is: does it answer questions directly, say who wrote it and when, and describe itself in the structured format AI systems read? A high GEO score means that when an AI assistant looks at your page, nothing stops it from using you as a source. (To see whether AI assistants actually mention you today, use AI Visibility — GEO score is the lever, AI Visibility is the scoreboard. Unfamiliar terms are defined in the glossary.)

Why GEO matters

When a user asks ChatGPT "What is the best tool for X?" or queries Perplexity for a comparison, the AI does not rank pages — it selects and synthesizes from sources it can understand quickly and trust. Pages that win citations in AI answers share common traits: clear structure, authoritative signals, rich schema, and factual density.

GEO is not about keyword stuffing or gaming algorithms. It is about making content machine-readable, trustworthy, and comprehensive enough that an AI can confidently cite it.

Schema markup (25 pts)

Schema is the single highest-weighted dimension. AI engines use Schema.org JSON-LD to understand what a page is (an Article, a Product, a How-To, a FAQ) and to extract structured facts directly.

Condition	Points
At least one JSON-LD block present	8
Block is valid parseable JSON with @context and @type	5
@type is a recognized, high-value Schema.org type	5
Rich properties present (author, datePublished, etc.)	4
Multiple complementary types (e.g. Article + BreadcrumbList)	3

High-value schema types for AI search

Type	Use case	AI benefit
Article / NewsArticle	Blog posts, news	Author, date, headline extracted
FAQPage	Q&A content	Direct answer extraction
HowTo	Step-by-step guides	Structured step synthesis
Product	Product pages	Price, availability, reviews
Organization / LocalBusiness	Brand/location pages	Entity recognition
BreadcrumbList	Navigation context	Page hierarchy for context
Review / AggregateRating	Reviews	Trust signal
Event	Events	Date, location extraction

Content depth (15 pts)

Condition	Points
Word count ≥ 600 words	4
Named entities detected (people, places, organizations)	3
At least one FAQ or How-To section present	4
Factual claims density (numbers, dates, specifics)	4

Citability signals (15 pts)

AI engines prefer sources they can trust and attribute. These signals indicate that a page is from a credible, maintained source.

Condition	Points
Author name present (byline or author schema)	4
Publish date present (datePublished or visible date)	4
Last modified date present (dateModified)	3
Organization or Person schema with name and url	4

Content structure (15 pts)

Well-structured content is easier for AI models to parse and cite selectively.

Condition	Points
Logical H2/H3 heading hierarchy (no skipped levels)	5
Answer-first paragraph structure (lead with conclusion)	4
Ordered or unordered lists present	3
Short introductory paragraph (≤ 150 words)	3

Meta & discoverability (10 pts)

Condition	Points
Title tag present and 30–60 chars	3
Meta description present and 70–160 chars	3
Canonical URL self-referencing	2
OG tags (og:title, og:description, og:image) all present	2

Citation patterns (10 pts)

Condition	Points
At least 2 external links to authoritative domains	4
External links have descriptive anchor text	3
Internal linking to related content (≥ 3 internal links)	3

GEO lint rules (10 pts)

The lint pass checks 35 writing patterns that reduce AI citability. Each failing rule subtracts from the 10-point pool.

View all 35 lint rules

Passive voice patterns — AI models prefer active, direct sentences

• Overuse of passive voice (> 20% of sentences)
• "It is believed that", "It has been said that"
• "was found to be", "is considered to be"

Hedge words — reduce perceived authority

• "might", "possibly", "arguably" as sentence openers
• "some people think", "many experts believe" without attribution
• "it depends", "it varies" without qualification

Vague openers — delay the answer

• "In today's world", "In this day and age"
• "Throughout history", "Since the beginning of time"
• "There are many reasons why"

Filler phrases — add words, subtract meaning

• "It is important to note that"
• "It goes without saying"
• "Needless to say"
• "As we all know"
• "At the end of the day"

Weak structure signals

• No paragraph breaks (wall of text)
• Paragraphs longer than 150 words
• No lists in content longer than 800 words
• Headings that don't describe content (e.g. "Introduction", "Overview" only)

Citation issues

• External links with generic anchor text ("click here", "learn more", "read more")
• Links to redirect chains
• No external links at all

Keyword and entity issues

• Primary topic not mentioned in first 100 words
• No named entities in first 300 words

Heading optimization (5 pts)

Condition	Points
At least 3 H2/H3 headings	2
At least one heading phrased as a question	2
Primary keyword in at least one heading	1

The 9 GEO improvement strategies

When SitePulse generates GEO recommendations, each recommendation maps to one of nine strategies. Understanding the strategy helps you batch fixes.

Strategy	What it addresses
cite_sources	Add citations to external authoritative sources
add_quotes	Add direct quotes from experts or primary sources
add_statistics	Include specific numbers, percentages, or data points
add_authority	Add author bio, credentials, or organization schema
fluency	Fix passive voice, hedge words, and sentence complexity
simpler_language	Reduce reading level, avoid jargon without explanation
unique_words	Increase vocabulary diversity and reduce repetition
add_keywords	Include topic-relevant terms in headings and early content
ease_of_understanding	Improve structure, add lists, break up dense paragraphs

How to improve

GEO improvements cluster into three effort levels:

Low effort (HTML changes):

• Add JSON-LD FAQPage or HowTo blocks to existing content
• Add datePublished, dateModified, author to your Article schema
• Rewrite hedge sentences to active voice

Medium effort (content rewriting):

• Add an FAQ section to high-traffic pages
• Add a "Last updated" date and author byline
• Restructure paragraphs to answer-first order

High effort (content expansion):

• Expand thin content (< 600 words) with supporting detail
• Add external citations with descriptive anchor text
• Add statistics and named entities

The Rewrites tab in your audit generates ready-to-apply schema blocks, FAQ sections, and citation suggestions. The cite_sources and add_authority strategies have the highest average impact on GEO score.