How SearchScore Calculates Your AI Citation Probability
SearchScore analyses any website using a 3-layer model, 23 scoring categories and 250+ individual signals, returning a single score from 0 to 100 across each of the three disciplines: AI Visibility (GEO), Search Engine Optimisation (SEO), and Conversion Rate Optimisation (CRO).
This page explains exactly what we check and why each signal matters.
This page describes the per-site SearchScore audit. For the State of AI Visibility Index (SAVI) (the periodic benchmark across 850,000+ sites), see searchscore.io/savi.
What the Three Audit Layers Measure
AI engines go through three stages when deciding whether to cite your site. Our GEO scoring model mirrors this process.
Technical basics, schema markup, crawl access, AI bot permissions and knowledge graph signals. Covers what Google's official AI Optimisation Guide recommends for AI Overviews and AI Mode visibility. If this layer is weak, AI may not even see your site.
Original data, query matching, topic depth, FAQs, direct answers, content freshness, author bios and citable passage structure. If this layer is weak, AI won't quote your content in answers.
Press coverage, reviews, community presence, backlinks, partner links, NAP consistency and social proof. If this layer is weak, AI won't recommend you to users.
Layer scores are conceptual groupings that help you understand where to focus. Your overall score (0-100) is the number that matters most.
The Three Disciplines and Their Scoring Categories
SearchScore audits 23 categories across three disciplines. Each returns a score from 0 to 100.
GEO Score - AI Search Visibility (8 categories)
Your GEO Score measures how likely AI engines - ChatGPT, Perplexity, Claude, Google AI Overviews - are to discover, understand, and cite your site.
| Category | What it measures |
|---|---|
| AI Citability | Can AI extract citable answers from your content? Checks FAQ sections, answer-first formatting, direct definitions, original data points and content freshness. |
| Brand Authority | Does your brand exist as a recognised entity online? Press coverage, Wikipedia/Wikidata presence, YouTube, podcasts, backlinks, community mentions and partner links. |
| EEAT Content | Does your content demonstrate genuine expertise? Named authors, bylines, credentials, expert endorsements, comparison pages, buyer decision guidance and use-case coverage. |
| Structured Data | Is JSON-LD schema markup present and correctly implemented? Checks Organisation, Article, Person, FAQ, HowTo, Speakable and brand entity definition. |
| Technical SEO | Are the technical fundamentals in place? HTTPS, crawlability, mobile performance, page rendering, canonical tags, indexability and crawl efficiency. |
| Platform Optimisation | Local business signals: name/address/phone consistency across directories, Google Business Profile, reviews and testimonials. |
| Topical Authority | Does your site demonstrate depth on its subject? Topic cluster structure, semantic breadth, internal linking and content hub signals. |
| AI Platform Readiness | Are AI crawlers explicitly permitted? Checks Google (Googlebot, Google-Extended for Gemini), Bing (Bingbot, IndexNow), and independent AI platforms (GPTBot, ClaudeBot, PerplexityBot, GrqBot, DeepSeekBot) in robots.txt. Also checks for AI-friendly structured data (FAQ, HowTo, QA). |
SEO Score - Search Engine Optimisation (8 categories)
Your SEO Score measures how well your site performs across traditional search ranking signals.
Category weights are not equal - categories marked with higher weight have more influence on your overall SEO score.
| Category | Weight | What it measures |
|---|---|---|
| Technical | High | HTTPS, www consistency, canonical, noindex detection (meta + X-Robots-Tag), sitemap, robots.txt quality, hreflang, internal links, security headers. |
| On-Page | High | Title tag, meta description, Open Graph, Twitter Card, breadcrumbs, H1 count and keyword alignment, image alt text (meaningful images), WebP images, schema types, duplicate scripts. |
| Content | High | Word count (scripts stripped), paragraph quality, content freshness signals, primary intent match. |
| AI / Crawler | High | AI bot access in robots.txt (multi-agent stanza handling, Crawl-delay detection), llms.txt, FAQ sections, lists, answer-first content, rich schema. |
| User Experience | Standard | Mobile viewport, language attribute, skip links, lazy loading, font-display swap, accessible click targets. |
| Topical Authority | Standard | H2/H3 heading structure, content depth (500+ words), internal links, question headings, content hub signals. |
| Authority | Standard | Outbound links, LinkedIn presence, social profile links. |
| E-E-A-T | Standard | Named author, bio, contact information, About page, privacy policy, credentials, press mentions. |
CRO Score - Conversion Rate Optimisation (7 categories)
Your CRO Score measures how well your site converts visitors into leads and customers.
| Category | Max pts | What it measures |
|---|---|---|
| Trust & Credibility | 36 | Owner/About credibility, real vs stock photos, reviews and testimonials, before/afters, social proof (client logos, press mentions, awards). |
| Above the Fold | 25 | H1 benefit clarity, subheadline quality, primary CTA visibility, value proposition clarity, social proof above fold. |
| Navigation & UX | 21 | Navigation labels, mobile responsiveness, fast load signals, design consistency, accessibility (ARIA landmarks, skip links, lang attribute). |
| Lead Capture & Funnels | 28 | CTA button presence and quality, contact forms, live chat, WhatsApp, multiple conversion touchpoints, form experience (field count, labels, error handling), CTA diversity (mid-page, sticky, footer). |
| Strategy & SEO | 41 | Niche-specific messaging, outbound readiness, local SEO, image accessibility, H1 structure, Open Graph, performance signals (LCP, render-blocking scripts, WebP, preconnect), pricing transparency. |
| Authority & Trust | 24 | Social media profile links (verified), press/media mentions, certifications and accreditations, trust badges. |
| LLM Optimisation | 24 | FAQ/Q&A sections, structured data (JSON-LD), quotable statistics and data points, clear content structure with headings. |
CRO scores are normalised to 0-100. The raw maximum is 199 points across all 7 categories.
What Each Score Tier Means
All three scores use the same tier system.
| Tier | Score Range | What It Means |
|---|---|---|
| Excellent / AI-Ready | 90-100 | Top performance. Fully optimised for AI citation, search ranking, and conversion. |
| Strong | 75-89 | Strong performance. Likely visible to AI and converting well in this discipline. |
| Competitive / Above average | 65-74 | Good fundamentals. Targeted improvements will close remaining gaps. |
| Emerging / Average | 50-64 | Visible for some queries. Significant opportunity with prioritised fixes. |
| Low / Below average | 30-49 | Multiple foundational issues. Structured improvement plan needed. |
| Critical / Invisible | 0-29 | Immediate action required. Core signals are missing or broken. |
The average across 850,000+ audited sites is 34/100 for AI visibility.
How We Collect and Verify Audit Data
Live page fetch
We fetch your page in real time using a standard browser user agent (Puppeteer/Chromium) and analyse the fully rendered HTML - headings, schema markup, meta tags, content structure, links, images and inline scripts.
robots.txt and sitemap
We fetch your robots.txt directly and check whether AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, GrqBot, DeepSeekBot) are permitted, including multi-agent stanza handling and Crawl-delay detection. We also check whether /sitemap.xml exists directly in addition to checking the Sitemap: declaration in robots.txt.
llms.txt
We fetch /llms.txt directly from your domain root and verify its presence. A page link to llms.txt is used as a fallback if the direct fetch fails. Note: llms.txt helps ChatGPT, Claude and Perplexity discover your content, but Google does not use it - Google relies on standard crawling and structured data instead.
External authority signals
We check for your brand's presence on Wikipedia, Wikidata, social platforms, Reddit, GitHub and Crunchbase. Social profile links found in your HTML are verified with live HTTP requests to confirm the profiles exist.
External reviews
We check Trustpilot, Google Reviews, Facebook and Yelp for review presence, ratings and review counts.
Performance signals
Where a full browser session is available, we measure Largest Contentful Paint (LCP) via the PerformanceObserver API. We also check for render-blocking scripts, WebP image usage, preconnect hints and lazy loading as performance proxies.
Conservative scoring
If a signal cannot be verified, we score it conservatively. We never assume a feature is present unless we can confirm it from the live page, HTTP response headers, or a direct resource fetch.
Aligned with Google's Official AI Optimisation Guide
Google published their official guide to optimising for AI Overviews and AI Mode in June 2025. Our GEO audit categories map directly to their recommendations:
| Google's Recommendation | Our Audit Category |
|---|---|
| "Create valuable, non-commodity content" | Content & E-E-A-T - unique POV detection, original data signals, content freshness |
| "Build and maintain a clear technical structure" | Technical SEO - HTTPS, crawlability, canonicals, sitemap, SSR, semantic HTML |
| "Add high-quality images and video" | Technical + Structured Data - image dimensions, responsive images, alt text, VideoObject schema |
| "Structured data" | Structured Data - JSON-LD, Organisation, Article, FAQ, VideoObject, Product schema |
| "Ensure crawlability" | AI Citability - robots.txt, AI bot permissions, crawler access |
| "Optimize local business and ecommerce" | Brand Authority - NAP consistency, Google Business indicators, local schema |
| "Reduce duplicate content" | Technical SEO - domain consistency, trailing slash handling, clean canonicals |
| "Good page experience" | Technical SEO - mobile viewport, accessibility, performance hints |
We also check signals that Google says are not required but benefit visibility across other AI platforms:
- llms.txt - not used by Google, but helps ChatGPT, Claude and Perplexity discover your content
- PerplexityBot access - specific to Perplexity, not Google
Source: Google's Guide to Optimizing for Generative AI Features on Google Search (last updated 2026-06-05)
What You Get Free
The free audit shows your overall score and tier for all three disciplines (GEO, SEO, CRO), plus the full category breakdown - 23 categories across the three audits - so you can see exactly where you stand.
The email-gated fix list unlocks the full prioritised action plan: every failing signal with a specific fix, urgency rating, and estimated time to implement.
Paid implementation - the AI Visibility Audit (custom pricing) and Monthly Growth Service (custom monthly engagement) - is where we do the work for you. We ship the fixes across all three disciplines and guarantee your GEO, SEO and CRO scores all improve within 30 days, or we work for free until they do.
How Often Scores Update
Free scans are cached for 7 days. If you have made significant changes to your site, you can request a fresh audit after 7 days. Paid users can force a rescan at any time.
The score reflects your site as it exists at the time of the audit - not how it was configured in the past. The audit timestamp is shown on all results pages so you always know how current the data is.
Track these scores weekly
The free audit is a snapshot. SearchScore Tracker runs weekly (or daily) scans across ChatGPT, Claude, Gemini, Perplexity, Grok and DeepSeek, telling you which AI engines cited you, where you ranked, and which competitors got mentioned instead. From $25/mo.
See Tracker plans →Run Your Free Audit
Free for any website. No email required. Results in under 60 seconds.
Get Your Free Score →