How Citaition monitors AI visibility

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.

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.

Every query is sent to each enabled AI provider – ChatGPT (OpenAI), Gemini (Google), Perplexity, Google AI Overviews, and Google AI Mode – 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.

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.

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 benchmark dataset.

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.

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 white-label reports coming soon. XLSX data exports provide raw data access for custom analysis. Reports are generated on your configured schedule – monthly, weekly, or daily.