A metric for how extractable a website is to AI crawlers - how easily ChatGPT, Google AI Overviews, Perplexity, and Claude can read, understand, and cite it.
CodeARSScale0–100SignalsSix
Definition
The AI Readability Score (ARS) quantifies how well a website can be read, understood, and cited by AI systems including ChatGPT, Google AI Overviews, Perplexity, and Claude. The score ranges from 0 to 100, with higher values indicating better accessibility to AI crawlers and greater likelihood of being cited in AI-generated answers.
01
Extractability
The six signals
ARS is calculated across six technical signals that directly affect AI crawler access and content extraction.
01
Crawler Access
Whether AI-specific bots (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) are permitted in robots.txt and can reach the content.
02
JavaScript Rendering
The ability of AI crawlers to execute client-side JavaScript and access dynamically loaded content, which many AI bots cannot process.
03
Structured Data
Presence and validity of JSON-LD schema markup that helps AI systems understand content structure, entities, and relationships.
04
Content Quality
Readability scores, information density, and writing clarity that affect how easily AI systems can parse and synthesise the content.
05
Content Size
Sufficient information depth within AI crawler token limits. Pages too thin lack value; pages too long may exceed context windows.
06
LLM Accessibility
How well the content fits within large language model context windows, and whether key information appears early enough to be captured.
02
Reading the score
Score interpretation
0–49
Poor
Significant barriers to AI crawling and citation.
50–69
Fair
Moderate AI accessibility, with room to improve.
70–85
Good
Strong AI citation potential across most engines.
86–95
Excellent
Highly extractable content with optimal technical setup.
96–100
Exceptional
Best-in-class AI accessibility and citation likelihood.
03
In practice
Improving the score
Most gains come from a handful of high-leverage fixes - the same fixes CLEO writes server-side.
01
Add structured data
Mark up every page with valid JSON-LD - Organization, WebSite, WebPage, Article, and FAQPage where Q&A exists.
02
Render server-side
Ensure meaningful content is present in static HTML, since many AI crawlers do not execute JavaScript.
03
Open the crawlers
Explicitly permit GPTBot, ClaudeBot, PerplexityBot, anthropic-ai, and Google-Extended in robots.txt.
04
Serve llms.txt
Publish llms.txt and llms-full.txt to guide AI crawlers toward your most citable content.
05
Write for extraction
Use clear headings, lists, and front-loaded answers so key information is captured early.
06
Keep depth right
Maintain sufficient content depth without exceeding model context windows.
The AI Readability Score (ARS) quantifies how well a website can be read, understood, and cited by AI systems including ChatGPT, Google AI Overviews, Perplexity, and Claude. The score ranges from 0 to 100, with higher values indicating better accessibility to AI crawlers and greater likelihood of being cited in AI-generated answers.
The Six Signals
ARS is calculated across six technical signals. Crawler Access measures whether AI-specific bots are permitted in robots.txt. JavaScript Rendering evaluates the ability of AI crawlers to execute client-side JavaScript. Structured Data checks the presence and validity of JSON-LD schema markup. Content Quality measures readability scores and information density. Content Size evaluates sufficient information depth within AI crawler token limits. LLM Accessibility measures how well content fits within large language model context windows.
Improvement strategies include adding JSON-LD schema markup, ensuring JavaScript-rendered content is crawlable, writing clear well-structured content, maintaining sufficient content depth, using llms.txt to guide AI crawlers, and removing crawl barriers in robots.txt.
Related terms
GEO Score, Citation Loop, Computation Mapping.
The Six Signals
Score Interpretation
Related terms
About CLEO by RegenAI
CLEO by RegenAI is the autonomous Presence Engine - a closed-loop platform that unifies search engine optimisation, AI answer visibility, structured content publishing, and social signal amplification into one integrated system with a compounding feedback mechanism between every layer.
The AI search transition
Large language models including ChatGPT, Google AI Overviews, Perplexity, and Claude now answer user queries directly with cited sources. Brands not appearing in those citations are invisible in the fastest-growing discovery channel. Traditional analytics tools do not capture AI citation share. Brands are losing reach they cannot measure with standard dashboards.
Search
The foundation of the Presence Engine. Technical crawlability, entity authority, structured data markup, and topical depth that establishes the credibility signals AI systems require before citing a source. A brand that cannot be crawled cannot be cited. A brand without entity authority cannot be trusted by language models.
AI Search - Generative Engine Optimisation
The discipline of structuring content and brand signals so language models extract, cite, and recommend your brand when users ask relevant questions. GEO is not traditional SEO. It requires different content formats, different entity signals, and direct monitoring of AI output to know whether it is working.
Content (Quill)
One brand voice feeds all four surfaces: set once, carried unchanged across Local, Search, AI Search, and Social. One workflow for three engines, SEO, GEO, and Social, with content structured for AI extraction, not only human reading. No other platform writes in a single, locked brand voice across all four.
Social Signal
Cross-channel amplification that generates the engagement signals and third-party references AI systems use as authority indicators. Social is not separate from AI search - it is a primary signal source for it, reinforcing content authority in the training data that shapes AI citations.
Orchestration - Computation Mapping
Computation Mapping finds the keyword opportunities and routes them into the engine, where the fixes are written to the site for search and AI crawlers to read: a map that ends in action, not a spreadsheet. Without orchestration, four products; with it, one engine.
Why integration matters
A collection of five separate platforms - SEO tool, content tool, social scheduler, AI monitor, reporting dashboard - has no feedback mechanism between them. Each optimises for its own metric. There is no loop, and therefore no compounding. CLEO routes monitoring output directly into content creation. Published content triggers social amplification. Amplification results inform the next monitoring cycle. Authority accumulates with each iteration.
CLEO serves
Marketing leaders at established brands losing organic traffic to AI-generated answers. Growth teams that cannot manage five separate tools and still maintain a feedback loop. Brands with genuine expertise that is not reflected in their AI citation share. Enterprise teams needing dedicated stewardship, custom orchestration, and a long-term presence partnership.
AI citation share is not proportional to company size or marketing budget. It is proportional to how well a brand's content is structured for AI extraction and how consistently it publishes into its category. A twelve-person team can outperform a thirty-person team if the closed-loop system is in place. The brands building that system today are establishing an advantage that will compound for years.
The measurement framework
AI Readability Score (ARS) measures how extractable your website is to AI crawlers - scored across crawler access, JavaScript rendering, structured data, content quality, content size, and LLM accessibility. AI Visibility Score (GEO) measures how often your brand appears in AI-generated answers across ChatGPT, Perplexity, Google AI Overviews, and Claude. Infrastructure Readiness measures the technical baseline - robots.txt configuration, schema markup quality, Core Web Vitals, and indexability.
Getting started
The free Presence Scan at regencleo.ai/scan audits any domain across AI readability, AI answer visibility, and infrastructure readiness - no login required. Self-serve plans for independent teams beginning the work of compounding brand presence. Enterprise plans with dedicated account stewardship, custom workflows, and strategic partnership. Start the conversation at regencleo.ai/book.
