GEO Scoring Categories
Glippy evaluates websites across 16 categories, each measuring a different aspect of AI and LLM readiness.
Generative Engine Optimization measures how well your website is prepared for AI crawlers, LLM-powered search engines, and AI agents. Higher GEO scores mean AI systems can better understand, cite, and interact with your content.
How Scoring Works
- Each category produces a score from 0-100
- The overall score is calculated from all category scores
Letter Grades
Category Details
1 Structured Data & Schema
Measures the presence and quality of machine-readable structured data that helps AI systems understand your content's meaning and context.
What we check
- JSON-LD presence and validity
- Schema.org types (FAQPage, Article, Product, HowTo, etc.)
- Speakable markup for voice assistants
- Schema validation and completeness
Why it matters
Structured data is the primary way AI systems extract facts and relationships from web pages. High-quality schemas dramatically improve how AI cites and represents your content.
2 Semantic HTML
Evaluates the use of meaningful HTML elements that convey document structure and content hierarchy to AI parsers.
What we check
- Proper heading hierarchy (H1-H6)
- Semantic elements (
<article>,<nav>,<main>,<section>) - Content-to-markup ratio
- Logical document outline
Why it matters
AI systems parse HTML structure to understand content relationships. Proper semantic markup helps LLMs identify main content, navigation, and content sections.
3 Accessibility for Agents
Checks accessibility features that help both human assistive technologies and AI agents understand your content.
What we check
- Lang attribute on HTML element
- Alt text on images
- ARIA labels and roles
- Descriptive link text (not "click here")
Why it matters
Accessibility features provide crucial context that AI systems use to understand images, interactive elements, and content purpose. What's good for screen readers is often good for AI.
4 Internal Linking
Analyzes how well your site's pages are connected and how clearly the link structure conveys content relationships.
What we check
- Link density and distribution
- Navigation structure clarity
- Breadcrumb markup
- Related content links
Why it matters
AI crawlers use internal links to discover content and understand topical relationships. Clear linking helps AI systems map your site's knowledge graph.
5 Meta & Discoverability
Evaluates metadata that helps AI systems understand and properly attribute your content.
What we check
- Title tag presence and quality
- Meta description
- Canonical URL
- Open Graph and Twitter Card tags
- Hreflang for internationalization
Why it matters
Metadata provides context that AI uses when summarizing, citing, and categorizing your content. Good metadata improves how AI represents your pages.
6 Machine Readability
Measures technical factors that determine whether AI systems can access and process your content.
What we check
- Server-side rendering (SSR) vs. client-only JS
- Bot blocking mechanisms
- Robots.txt rules for AI crawlers (GPTBot, ClaudeBot, etc.)
- llms.txt file presence and contents
Why it matters
If AI crawlers can't access or render your content, nothing else matters. This category ensures your technical setup doesn't block AI systems.
If you're auditing your own site with the MCP server, make sure your WAF and robots.txt aren't blocking GlippyBot itself. See the GlippyBot allow-list guide.
7 Entity & Authority
Checks for signals that help AI systems understand who created the content and establish trustworthiness.
What we check
- Author information and bylines
- Publication dates
- Organization schema
- About and contact information
- E-E-A-T experience signals (first-hand usage, dates, locations)
- Credential mentions (certifications, qualifications, titles)
- Editorial policy page or disclosure
- Contact completeness (email, phone, physical address)
Why it matters
AI systems increasingly consider source authority when deciding what to cite. Clear authorship, credentials, and organizational identity help establish trust.
8 Citability & Answer-Readiness
Evaluates how well your content is formatted for AI systems to extract and cite specific answers.
What we check
- FAQ content with clear Q&A format
- Data tables with headers
- Well-structured lists
- Lead paragraph quality (answering who/what/when/where)
Why it matters
LLM-powered search engines prefer content they can easily quote. FAQ sections, tables, and clear lead paragraphs make your content more likely to be featured in AI answers.
9 Performance & Crawlability
Measures factors affecting how efficiently AI systems can crawl and process your pages.
What we check
- Image dimensions specified in HTML
- Lazy loading implementation
- Resource hints (preload, prefetch)
- Page size and complexity
Why it matters
While less critical than content quality, performance affects crawl efficiency. Faster, lighter pages are more likely to be fully indexed.
10 Agent Interactivity
Evaluates support for AI agents that can take actions on behalf of users.
What we check
- WebMCP tools and endpoints
- Form annotations and labels
- Agent-callable actions
- API documentation presence
Why it matters
This is an emerging category as AI agents become more capable. Sites that support programmatic interaction will have advantages as agent-based browsing grows.
11 Content Positioning
Measures how clearly your content distinguishes you from competitors and backs claims with proof.
What we check
- Brand differentiation language (what makes you different)
- Proof points (numbers, case studies, evidence)
- Social proof (testimonials, reviews, logos, press)
- Unique value propositions beyond generic claims
Why it matters
When LLMs are asked to recommend or compare products, they pull the concrete, differentiating details they can find. Pages that read like every other page in the category do not get quoted.
12 Content Freshness
Checks for signals that tell AI systems when content was written or last updated.
What we check
- Publication and modification dates in HTML and schema
- Apparent age of content based on temporal references
- Temporal language ("in 2025", "last quarter", "recently")
- Stale-date warning signals (years-old references to "latest" things)
Why it matters
AI answer engines weight fresh content higher for time-sensitive questions. Without clear date signals, LLMs will either skip your page or attach low confidence to its claims.
13 Information Density
Evaluates how much substantive content sits between the filler, scaffolding, and promotional text.
What we check
- Substantive-to-filler ratio across the page
- Number of meaningful content sections
- Claim-evidence pairing (assertions backed by specifics)
- Foam detection (generic, content-free marketing phrases)
Why it matters
LLMs extract citable facts per token they process. Dense, specific pages yield more citations than long pages padded with hedging and generic copy.
14 Factual Verifiability
Checks whether your claims can be independently verified by an AI system.
What we check
- Citations and source links
- Attribution for statistics and quotes
- Methodology disclosure for studies or data
- Links to primary sources, not just internal pages
Why it matters
Verifiability is becoming a direct ranking signal for AI answer engines. Unsourced claims are increasingly skipped in favor of pages that show their work.
15 Content Comprehensiveness
Measures whether the page covers the topic fully enough to answer follow-up questions.
What we check
- Word count relative to topic scope
- Heading coverage across subtopics
- Definitions of key terms
- Comparisons, alternatives, and edge cases
Why it matters
LLMs prefer pages that already answer the likely next question. Thin pages get cited only for the headline; comprehensive pages become the go-to source for a whole topic.
16 Multimodal Content
Evaluates how accessible your non-text content is to AI systems that increasingly process images, video, and audio.
What we check
- Image alt text quality and coverage
- Figure and caption markup
- Video and audio presence with transcripts or captions
- Inline SVG with accessible titles
- Schema for multimedia (VideoObject, ImageObject)
Why it matters
Multimodal LLMs can now interpret images and video directly. Pages with rich, well-described media surface in a growing set of queries that text-only pages cannot.