Live - tracking ChatGPT, Gemini & ClaudeNew - Google AI Overviews & DeepSeek on the roadmapFree audit - see how AI sees youEvery metric ships a 95% confidence range
Glossary

The language of AI search visibility.

AI search has its own vocabulary — fan-out, selectability, citability signatures. Here is the short version, in plain English.

Reference

Every term, one card.

Concept

GEO / AEO

Generative & Answer Engine Optimization: getting your brand cited inside AI-generated answers, the way SEO targets blue-link rankings.

Concept

Fan-out

Before answering, an engine rewrites your one prompt into roughly 8–9 targeted sub-queries. CitedOS tracks those sub-queries, not just the headline question.

Concept

Citation / source

A URL the engine referenced to ground its answer. Around 90–95% of citations point to third-party pages — reviews, forums, press — not your own site.

Method

Grounding

The engine's built-in web-search step that pulls live sources into an answer. We read each engine's grounding metadata to learn which sources it actually used.

Concept

Selectability

Once a page is in the retrieval pool, will it survive the rerank and get cited? Driven by on-page signals: passage structure, fact density, schema, freshness, speed.

Concept

Discoverability

Is the page even in the retrieval pool to begin with? Driven by off-site authority and search-index coverage — not by how the page itself reads.

Method

Citability signature

The learned set of on-page features that separate sources an engine cites from the ones it ignores for a given type of query.

Metric

Share of voice

The share of AI answers in your category that mention you, measured across multiple runs with a confidence range on every number.

Metric

Mention rate

How often you appear at all for a tracked prompt, regardless of rank or sentiment — the broadest on/off signal.

Metric

Position

Where you land in the answer when you are mentioned. Being named first carries more weight than a passing footnote.

Metric

Sentiment

Whether the answer frames you positively, neutrally, or negatively on the occasions it brings you up.

Metric

Presence

Whether you show up across the engines and prompts we track — the coverage view that sits beneath the finer metrics.

Honesty

Confidence range

AI citations swing 40–60% month to month, so every CitedOS number ships with a 95% range and a freshness date instead of one fragile point-in-time figure.

Honesty

Proxy surface

The ChatGPT API web-search tool approximates, but is not identical to, the consumer ChatGPT app. We label that data a proxy surface rather than overclaim it.

Method

RAG / CAG

Retrieval- vs Cache-Augmented Generation: how CitedOS pulls from a large source corpus (RAG) and reuses bounded per-brand context (CAG) to analyze citability affordably.

See these metrics on your own brand.