How To Calculate Aeo Score

An AEO score is a composite metric on a 0–100 scale estimating how ready, extractable, and visible a brand or its content is for AI answer engines. Unlike traditional SEO scores—which prioritize keyword rankings, backlink authority, and click-through rates on Google Search—AEO scores measure whether ChatGPT, Perplexity, Google AI Overviews, and Claude cite a domain when answering user queries. The core difference is audience: SEO targets human searchers clicking links; AEO targets LLM systems extracting and attributing passages.

AEO scores reflect three distinct layers:

• AEO prioritizes AI citation frequency and content extractability over search ranking position
• SEO focuses on human click-through rates and keyword visibility in search results
• AEO requires schema markup, answer-first content structure, and LLM-readable formatting
• SEO relies on backlinks, on-page keywords, and domain authority signals

A brand with high traditional SEO rankings but poor content structure for LLM extraction will score lower on AEO. Conversely, a brand with extractable, well-structured content may earn AI citations even without dominant Google rankings. Specifically, a B2B SaaS company using Fastlook's Page Engine to publish answer-first content with JSON-LD schema may achieve higher AEO scores than a competitor ranking #1 on Google but publishing dense, unstructured articles.

A brand-level AI visibility score is calculated by testing a set of target category questions and assigning points based on brand placement in AI responses. The method follows a standardized formula: divide total points earned by maximum possible points, then multiply by 100 to generate a 0–100 score. For example, if you test 20 queries related to your product category and earn 75 points out of a possible 100, your brand-level AI visibility score is 75.

Point allocation typically follows this structure:

• Test 15–30 category-relevant queries across 4+ AI engines
• Assign 5 points for featured placement, 3 for secondary, 1 for tertiary
• Formula: (points earned ÷ maximum points) × 100 = brand AI visibility score
• Re-test monthly to track citation trend and competitive positioning

Testing occurs across multiple AI engines—ChatGPT, Perplexity, Google AI Overviews, and Claude—to capture breadth of visibility. The calculation method ensures comparability across brands and time periods. Tracking this score monthly reveals whether your content strategy is increasing AI citation frequency or losing ground to competitors. This metric is most useful for B2B SaaS and D2C brands competing for visibility in category-level queries; for instance, testing "best project management software" or "how to calculate customer lifetime value" across all four engines provides a reliable benchmark.

On-page content extractability scores compute a weighted average of technical, structural, and semantic dimensions to measure how easily LLMs parse and cite your content. The three core dimensions are:

• Technical: JSON-LD schema, heading hierarchy, meta tags, robots directives
• Structural: answer-first opening sentences, 10–20 word sentences, native bullet lists
• Semantic: 3+ named entities per passage, inline citations, 1+ grounded fact per section

Each dimension receives a weighted score, typically 35% technical + 35% structural + 30% semantic = extractability score. A page with proper schema markup but dense paragraphs and no citations scores lower than a page with moderate markup but clear answer-first sections and three inline source links. According to Schema.org documentation, implementing FAQPage, Article, or NewsArticle markup with structured answer blocks significantly improves LLM extractability. For instance, a Fastlook Page Engine article with JSON-LD schema, 15-word sentences, and citations to industry standards will score higher on extractability than an unstructured competitor article. Tools that audit extractability typically flag missing schema, sentences exceeding 25 words, paragraphs without citations, and low entity density as high-priority fixes.

The primary AI answer engines evaluated in AEO scoring are ChatGPT (OpenAI), Perplexity, Google AI Overviews, and Claude (Anthropic). These four platforms account for the majority of AI-generated answers and represent distinct user bases and query patterns. ChatGPT dominates conversational search and long-form research; Perplexity specializes in real-time information and cited answers; Google AI Overviews appear directly in Google Search results; Claude serves enterprise and technical audiences. Testing all four ensures AEO scores reflect real-world visibility across the AI search landscape.

• ChatGPT: conversational, long-form, broad audience
• Perplexity: real-time, cited answers, research-focused
• Google AI Overviews: search-integrated, launched May 2024
• Claude: enterprise, technical, reasoning-heavy queries
• Track crawler visits: GPTBot, ClaudeBot, PerplexityBot, Googlebot-Extended

Some brands also track secondary platforms—Grok (X), Microsoft Copilot, and niche vertical engines—depending on audience overlap. However, the calculation weights each engine equally unless audience data shows skewed platform usage. For instance, if 70% of target users query Claude, weight Claude at 40% and distribute the remaining 60% across the other three engines. Tracking AI crawler visits via server logs confirms which engines are actively indexing content. A complete AEO audit includes citation frequency across all four major engines plus crawler access logs to validate that pages are discoverable.

Several frameworks and tools exist for auditing and tracking AEO scores, each using slightly different methodologies but converging on the 0–100 scale. Manual audit frameworks require testing 15–30 target queries across 4 AI engines, recording placement (featured, secondary, tertiary), and calculating the brand-level score using the formula above. Automated audit tools scan your live pages for schema markup completeness, heading structure, sentence length, entity density, and citation presence, then generate an on-page extractability score.

• Automated on-page audits identify technical and structural gaps
• Manual query testing measures actual citation performance
• Severity-weighted issue flags: critical, high, medium, low
• Monthly re-audits track score trends and competitive positioning

Some tools also integrate AI crawler logs to show which pages are being indexed and how frequently. The most comprehensive approach combines both methods: automated on-page audits (to identify technical and structural gaps) plus manual query testing (to measure actual citation performance). For instance, Fastlook's Site Audit generates severity-weighted issue reports and one-click fixes for the weakest pages. Tools typically flag issues by severity—critical (missing schema), high (sentences >25 words), medium (low entity density)—and suggest fixes. Version control matters: tools that re-audit monthly and track score trends over time reveal whether content improvements are translating to higher AI visibility. A complete audit framework should cover all three dimensions (technical, structural, semantic) and test across all four major AI engines to avoid blind spots in one platform or content type.