How Citaition monitors AI visibility
From adding a brand to receiving actionable recommendations — a complete view of the pipeline that turns raw AI responses into intelligence your agency can act on.
Add a brand
Enter your client's brand name, website URL, category, and up to three competitors. Citaition immediately scrapes up to 100 pages from the brand's website — including blog posts, documentation, community forums, and help centres — extracting content depth signals, page structure, and word counts for every page. This website signal analysis forms the foundation for understanding what content AI assistants have to work with.
Design queries
Citaition generates approximately 100 strategically designed queries for the brand, covering the full buying journey. Category-level queries ('what is the best project management tool for agencies'), comparison queries ('Monday vs Asana vs ClickUp'), feature queries ('project management with time tracking'), and use-case queries ('project management for remote teams'). Each query is designed to mirror how real people ask AI assistants for recommendations in the brand's category.
Query AI assistants
Every query is sent to each enabled AI provider — ChatGPT (OpenAI), Gemini (Google), and Perplexity — with web search enabled. This is deliberate: web-search-enabled responses reflect the AI's real-time understanding based on current web content, which is the channel you can directly influence through content strategy. Each response is captured in full, along with every citation URL the AI used as a source.
Analyse responses
Every AI response is analysed to extract structured data: whether the brand was mentioned, its position in the recommendation list, sentiment (positive, neutral, or negative), whether a clear winner was identified, and which specific URLs were cited as sources. Citation URLs are classified as brand domain, competitor domain, or third-party source. This response-level analysis produces the raw data that drives every metric in the platform.
Compute diagnostics
Raw response data is aggregated into six diagnostic dimensions: training data presence (do AI models know the brand from their training data?), citation footprint (how much citable content exists on the brand's site?), content depth (are individual pages substantive enough to be useful sources?), competitive pressure (how much do competitors dominate the citation landscape?), recommendation strength (how strongly does AI endorse the brand when it does mention it?), and positioning clarity (does the brand show up consistently across its category?). Each dimension produces a 0-100 score calibrated against our 53-brand benchmark dataset.
Generate opportunities
The opportunity engine applies conditional rules to each brand's diagnostic profile, generating prioritised recommendations based on the brand's specific situation. A brand with low citation footprint and high bare-domain ratio gets content creation recommendations. A brand with high competitive pressure and low survival rate gets comparison content strategies. A brand with high search lift and low training data presence gets urgency-window guidance. Each opportunity includes evidence, a priority level, and a concrete action plan.
Generate report
A comprehensive PDF report is generated automatically, including headline visibility scores, provider-by-provider breakdowns, competitive intelligence, citation source analysis, content gap identification, and the full set of prioritised recommendations. Agency Pro subscribers receive white-label reports with their own branding. XLSX data exports provide raw data access for custom analysis. Reports are generated on your configured schedule — monthly, weekly, or daily.
Add a brand
Enter your client's brand name, website URL, category, and up to three competitors. Citaition immediately scrapes up to 100 pages from the brand's website — including blog posts, documentation, community forums, and help centres — extracting content depth signals, page structure, and word counts for every page. This website signal analysis forms the foundation for understanding what content AI assistants have to work with.
Design queries
Citaition generates approximately 100 strategically designed queries for the brand, covering the full buying journey. Category-level queries ('what is the best project management tool for agencies'), comparison queries ('Monday vs Asana vs ClickUp'), feature queries ('project management with time tracking'), and use-case queries ('project management for remote teams'). Each query is designed to mirror how real people ask AI assistants for recommendations in the brand's category.
Query AI assistants
Every query is sent to each enabled AI provider — ChatGPT (OpenAI), Gemini (Google), and Perplexity — with web search enabled. This is deliberate: web-search-enabled responses reflect the AI's real-time understanding based on current web content, which is the channel you can directly influence through content strategy. Each response is captured in full, along with every citation URL the AI used as a source.
Analyse responses
Every AI response is analysed to extract structured data: whether the brand was mentioned, its position in the recommendation list, sentiment (positive, neutral, or negative), whether a clear winner was identified, and which specific URLs were cited as sources. Citation URLs are classified as brand domain, competitor domain, or third-party source. This response-level analysis produces the raw data that drives every metric in the platform.
Compute diagnostics
Raw response data is aggregated into six diagnostic dimensions: training data presence (do AI models know the brand from their training data?), citation footprint (how much citable content exists on the brand's site?), content depth (are individual pages substantive enough to be useful sources?), competitive pressure (how much do competitors dominate the citation landscape?), recommendation strength (how strongly does AI endorse the brand when it does mention it?), and positioning clarity (does the brand show up consistently across its category?). Each dimension produces a 0-100 score calibrated against our 53-brand benchmark dataset.
Generate opportunities
The opportunity engine applies conditional rules to each brand's diagnostic profile, generating prioritised recommendations based on the brand's specific situation. A brand with low citation footprint and high bare-domain ratio gets content creation recommendations. A brand with high competitive pressure and low survival rate gets comparison content strategies. A brand with high search lift and low training data presence gets urgency-window guidance. Each opportunity includes evidence, a priority level, and a concrete action plan.
Generate report
A comprehensive PDF report is generated automatically, including headline visibility scores, provider-by-provider breakdowns, competitive intelligence, citation source analysis, content gap identification, and the full set of prioritised recommendations. Agency Pro subscribers receive white-label reports with their own branding. XLSX data exports provide raw data access for custom analysis. Reports are generated on your configured schedule — monthly, weekly, or daily.
The only channel you can influence today
AI assistants have two layers of knowledge: training data (a snapshot from months ago that you cannot influence today) and live web search (what they find right now). Citaition focuses on web-search-enabled responses because this is the channel where content strategy has immediate, measurable impact. Publish better content today and AI assistants can find and cite it tomorrow.
We do measure training data presence as a diagnostic dimension — and when combined with live citation visibility, it identifies brands in the urgency window: their content is working right now, but they haven't yet made it into the next training data update. What they do in the next 3-6 months determines whether they lock in that visibility permanently.
Ready to see how your clients appear in AI responses?
Add your first client brand and watch Citaition work through every step — from query generation to a complete AI visibility report with actionable recommendations.
No credit card required · Work email only