The CAVIS Framework

What is CAVIS?

CAVIS (Conversational AI Visibility Simulation) is a proprietary methodology for systematic AI visibility work, developed by Krister Ross and CitationLab AS. It is grounded in the analysis of 150,000 AI conversations from the WildChat dataset and provides a structured framework for understanding, measuring, and improving AI visibility.

The five CAVIS dimensions

C — Citation Rate

The share of relevant AI prompts in which the brand is mentioned. The primary KPI for AI visibility.

A — Authoritativeness

E-E-A-T profile and subject-matter authority — the prerequisite for AI models to evaluate content as a credible source.

V — Visibility Architecture

The technical infrastructure that makes content available to AI crawlers and machine-readable.

I — Information Quality

Content quality, answerability, and information gain — what actually determines whether content is cited.

S — Sentiment & Share

AI sentiment (positive/neutral/negative coverage) and share of voice relative to competitors.

Implementation

The CAVIS methodology can be implemented with CitationLab Monitor, which automates simulation, data collection, and reporting.

FAQ

What sets CAVIS apart from general AEO? CAVIS is a measurement and analysis framework. AEO is the broader discipline. CAVIS provides the structure for quantifying and systematically improving AEO work.

Is CAVIS free to use? The methodology is open and published. CitationLab Monitor automates CAVIS implementation.