How we measure AI visibility, the research that informs our scoring, and a transparent record of every refinement we have made to the methodology.
The way people find businesses is fundamentally changing. AI assistants like ChatGPT, Perplexity, and Google's AI Overviews are becoming the first point of contact for millions of searches daily.[4] Yet most businesses have no idea whether AI can find, understand, or recommend them.
We developed the CITE Score™ methodology to give business owners a clear, actionable picture of their AI visibility. Our research draws from multiple disciplines: information retrieval[1], natural language processing[11], local SEO[5], and Google's own quality guidelines.[2]
The goal is not just to measure — it is to provide a roadmap for improvement that any business owner can follow, regardless of technical expertise.
Our scoring system evaluates four critical dimensions of AI visibility. Each dimension is weighted equally (25 points) for a total possible score of 100. Every point is earned from verifiable, real-time data — never estimated or simulated.
Can AI systems find and read your website? We analyze schema markup[3], robots.txt configuration, page speed[12], HTTPS status, and XML sitemaps.[14]
Signal Breakdown:
| Schema presence | 0–5 pts |
| Breadcrumb schema | 0–3 pts |
| FAQ schema | 0–3 pts |
| PageSpeed score | 0–4 pts |
| Robots.txt AI access | 0–5 pts |
| HTTPS encryption | 0–3 pts |
| Sitemap + IndexNow | 0–4 pts |
Does AI know who you are? We check Organization and LocalBusiness schema[3], schema completeness (name, URL, logo, sameAs links, description), and entity signals.[8]
Signal Breakdown:
| Organization schema | 0–6 pts |
| LocalBusiness schema | 0–5 pts |
| Person schema | 0–2 pts |
| Name, URL, logo completeness | 0–5 pts |
| sameAs social links | 0–3 pts |
| Description present | 0–2 pts |
| Address + phone | 0–2 pts |
Does AI trust you? We evaluate backlink authority[9], brand search volume (the strongest single predictor of LLM citations, r=0.334)[4], review sentiment[7], and content recency.[11]
Signal Breakdown:
| Domain rank | 0–5 pts |
| Referring domains | 0–4 pts |
| Brand search volume | 0–6 pts |
| Dofollow ratio | 0–2 pts |
| Article + Product schema | 0–3 pts |
| Content recency | 0–5 pts |
| Review sentiment | 0–4 pts |
Are you where AI looks? We check Google Business Profile[5], multi-platform presence (sites on 4+ platforms are 2.8x more likely to be cited)[4], schema coverage, and content breadth.[6]
Signal Breakdown:
| Platform presence | 0–8 pts |
| Schema type coverage | 0–5 pts |
| LocalBusiness presence | 0–4 pts |
| GBP verification | 0–4 pts |
| Content breadth (FAQ + Article) | 0–4 pts |
Each CITE dimension is scored on a 0–25 scale based on multiple weighted factors. The total CITE Score™ is the sum of all four dimensions (0–100). Every data point is collected in real time from live website analysis, third-party APIs, and structured data extraction — no scores are estimated or simulated.
| Score Range | Label | What It Means |
|---|---|---|
| 80–100 | Excellent | AI systems actively recommend your business |
| 60–79 | Good | Solid foundation with clear room for improvement |
| 40–59 | Needs Work | AI sees you but does not yet trust you enough to recommend |
| 0–39 | Critical | Significant gaps in AI visibility need immediate attention |
We analyze up to 5 competitors you identify, plus we discover additional competitors through DataForSEO's competitive intelligence API.[10] Each competitor receives an estimated CITE Score based on the same methodology, allowing for direct comparison. When full schema data is unavailable for a competitor, we use a combination of DataForSEO traffic metrics and domain characteristics to produce a differentiated estimate. This helps identify specific areas where competitors outperform you and opportunities for differentiation.
Our audit system integrates multiple real-time data sources to ensure accuracy and completeness:
Schema.org Extraction
Live HTML parsing for JSON-LD structured data
Google PageSpeed Insights
Performance, accessibility, and Core Web Vitals
DataForSEO Backlinks API
Domain rank, referring domains, dofollow ratio
DataForSEO SERP API
Brand search volume and competitive intelligence
Google Places API
GBP verification, review data, and sentiment
Robots.txt Parser
AI bot access status for 8 major crawlers
Our methodology is grounded in established research from multiple fields. Each scoring signal maps to published findings about how AI systems discover, evaluate, and cite business information.
The foundational principle that links between pages signal authority was established by Brin and Page in their original PageRank research.[1] While modern AI systems use far more sophisticated methods, the core insight remains: external references (backlinks, citations, directory listings) are strong indicators of trustworthiness. Our Trustworthy dimension directly measures these signals through domain rank and referring domain analysis.[9]
Google's Experience, Expertise, Authoritativeness, and Trustworthiness (E-E-A-T) guidelines[2] inform our trust scoring across all four dimensions. We evaluate signals that demonstrate real-world expertise and authority in your industry, including structured data completeness, review sentiment, and content recency.
The emerging field of GEO[4] studies how content can be optimized specifically for AI-powered search engines. Key findings from this research inform our scoring: brand search volume is the strongest single predictor of LLM citations (r=0.334), multi-platform presence increases citation likelihood by 2.8x, and 65% of AI citations target content less than one year old. These findings directly shape our Trustworthy and Everywhere dimension weights.
For location-based businesses, we incorporate local SEO best practices from Whitespark's annual ranking factors survey[5] and Moz's state of local SEO report.[6] Google Business Profile signals, citation consistency, and review quality are among the top-ranked factors for local visibility — and these same signals influence how AI systems recommend local businesses.[7]
Recent surveys on how large language models process and retrieve information[11] inform our understanding of what signals AI systems prioritize when generating recommendations. Research on entity authority evaluation[13] helps us weight the relative importance of structured data, cross-platform consistency, and content quality.
A transparent record of every refinement to the CITE scoring system
Brin, S. & Page, L. (1998). The Anatomy of a Large-Scale Hypertextual Web Search Engine. Proceedings of the 7th International World Wide Web Conference.
Foundational work on link-based authority signals used in web ranking.
View sourceGoogle Search Quality Team (2024). Search Quality Evaluator Guidelines (E-E-A-T). Google.
Defines Experience, Expertise, Authoritativeness, and Trustworthiness as core quality signals.
View sourceSchema.org Community Group (2024). Schema.org Vocabulary Documentation. Schema.org.
The shared vocabulary for structured data markup used by Google, Bing, and AI systems.
View sourceAggarwal, P., Muralidhar, N., et al. (2024). GEO: Generative Engine Optimization. arXiv preprint arXiv:2311.09735.
First academic framework for optimizing content visibility in generative AI engines.
View sourceWhitespark (2023). Local Search Ranking Factors Survey. Whitespark Annual Report.
Annual survey of local SEO practitioners identifying top ranking factors including GBP signals and reviews.
View sourceMoz (2023). The State of Local SEO Industry Report. Moz.
Comprehensive analysis of local search ranking factors including citation consistency and review signals.
View sourceBrightLocal (2024). Local Consumer Review Survey. BrightLocal.
Annual consumer behavior study showing 87% of consumers read online reviews for local businesses.
View sourceGoogle Developers (2024). Structured Data General Guidelines. Google Search Central.
Official guidelines for implementing structured data to improve search visibility.
View sourceAhrefs (2024). Search Traffic Study: How Backlinks Correlate with Rankings. Ahrefs Blog.
Large-scale study demonstrating correlation between referring domains and organic search rankings.
View sourceDataForSEO (2024). SERP Features and AI Overviews Analysis. DataForSEO Research.
API-driven research platform providing competitive intelligence and SERP analysis data.
View sourceZhu, Y., Wang, R., et al. (2024). Large Language Models for Information Retrieval: A Survey. arXiv preprint arXiv:2308.07107.
Survey of how LLMs process and retrieve information, informing our understanding of AI citation behavior.
View sourceGoogle (2024). PageSpeed Insights & Core Web Vitals Documentation. Google Developers.
Performance metrics that influence both traditional and AI-powered search visibility.
View sourceOtterbacher, J., Bates, J., & Clough, P. (2024). Competent or Credible? How LLMs Evaluate Entity Authority. Proceedings of ACM SIGIR.
Research on how language models assess entity authority and trustworthiness signals.
View sourceMicrosoft Bing (2023). IndexNow Protocol Documentation. IndexNow.org.
Instant indexing protocol adopted by Bing, Yandex, and other search engines for faster content discovery.
View sourceWe continuously refine the CITE Score™ framework as AI search evolves. Opt in to receive an email whenever we publish a new methodology version.
Sign in to subscribe to methodology updates.
Sign In to Subscribe