Live, scanning the open webAI Visibility Audit Platform

The vocabulary of AI visibility

AI Visibility Glossary

The terms that describe how AI engines find, understand, trust, and recommend brands, defined in plain words.

AI visibility is a new field, and a new field needs shared language. These are the definitions AudFlo uses, and they are published as structured data so AI engines can read them too.

AI Visibility

How easily AI answer engines such as ChatGPT, Claude, Gemini, and Perplexity can find, understand, trust, and recommend a brand. It is not about ranking in a list of links. It is about being the clear, trusted, and easy to recommend answer for a given prompt.

The AI Visibility Guide

AI Recommendation

When an AI answer engine names a specific brand or product in its answer to a user, as a suggestion or a citation, rather than only returning a list of links. An AI recommendation is the outcome that AI visibility work is trying to earn.

How AI engines decide

AI Recommendation Gap

The distance between the brands an AI engine recommends and the brands that deserve to be recommended. A company has a recommendation gap when AI engines suggest competitors instead of it, usually because the AI cannot classify, verify, or trust it with the same confidence.

Why ChatGPT recommends competitors

Entity Confidence

The likelihood that an AI system can correctly identify who you are, what category you belong to, and when to recommend you. It is built through consistent naming, a single clear category, and structured data such as Organization and SoftwareApplication schema, so an AI can tell your brand apart from similar names.

The Entity Confidence framework

Contextual Fit

How well a brand matches the exact prompt a user types. Broad claims lose. Specific positioning, comparison content, and FAQ coverage tuned to the real questions buyers ask are what make a brand the answer to those prompts.

How it is measured

Evidence Ladder

A five-tier ranking of how much AI engines trust a proof signal, strongest first: original research and proprietary data, then independent third-party validation, then named case studies, then named testimonials, with unverified marketing claims at the bottom.

The Evidence Ladder in full

Category Clarity

How clearly a site states what category its product belongs to and who it serves, in plain words, in the places an AI reads first: the H1, the title, the metadata, and the opening copy. Clear category language is what lets an AI classify a product before it can recommend it.

Why startups are invisible to ChatGPT

AI Visibility Score

AudFlo's measure, from 0 to 100, of how well AI engines can understand, classify, and recommend a site. It is computed from 32 checks across four weighted pillars: Technical Visibility, Structural Understanding, Answer Selection, and Authority Signals.

How the score works

Answer Engine Optimization

The practice of making a site easy for AI answer engines to read, trust, cite, and recommend. It is to AI engines what search engine optimization is to traditional search results. It is often shortened to AEO.

The AEO Playbook

Recommendation Confidence

The overall likelihood that an AI engine will recommend a brand, modeled as the product of five factors: Accessibility, Classification, Evidence, Entity Confidence, and Contextual Fit. If any one factor is near zero, the whole result collapses.

The five factors

See these ideas in action in the AI Visibility Playground, go deeper in the Evidence Ladder, Entity Confidence, and Benchmark Framework, or run a free scan to see where your own site stands.