Glossary

The vocabulary, in the words the product uses. Precise definitions and windows are in data definitions.

AEO

Answer engine optimisation: shaping your site and content so AI engines have something to quote when they answer a question in your category. Not SEO under a new name. A top-ranking page can be quoted by nothing.

AI Overviews

The AI panel Google draws above its own search results. It is read from the results page rather than asked a question, and it is a different surface from Gemini. When Google draws no panel, the check is recorded as an absent overview and left out of your rate.

Arrival

A visit the Trace tag reported as coming from an AI assistant. Weaker than a bound visit.

Assistant

The front door of the product. You ask a question in your own words and it answers from your workspace's counted rows, opening the evidence beside the answer.

Attested

A change tier: a person approved it and says it shipped. Their word, recorded as their word, never treated as a measurement.

Bound visit

A Trace visit joined to a citation the platform counted. The only figure that proves an answer sent somebody.

Change

Something you shipped in response to what a lens showed, tracked through the ledger from proposed to confirmed.

Citation

Your business named in an answer as a genuine option, by brand, product, service, URL or phone number.

Citation rate

Citations divided by the checks that count, over a stated window.

Confidence score

How sure the models are that a citation is real and substantive. Not a quality score, not a ranking.

Confirmed moved

A change tier: the answers moved, scored against a control captured when the change shipped.

Coverage

The wallet of answers your plan allocates each period, split into monitoring and generative pools. Everything the platform does on your behalf spends from it.

Drifted

A change tier: it was observed live, and the page has since stopped satisfying it. Deliberately ranked above confirmed so a win that has come undone does not read as a clean one.

Engine

One of the six chat surfaces asked a question: ChatGPT, Claude, Gemini, Perplexity, Grok, DeepSeek. Together with AI Overviews they are the seven monitored surfaces.

Excluded

A surface with no measurement in the window, reported as an exclusion rather than as zero.

Intent

The category a prompt falls into: local, buying, comparison, informational or branded. Classified before anything is asked.

Judge

The model that reads every answer and decides whether you are genuinely named.

Landscape, subject, proof

The three depths every lens descends: everything at once, one row opened, and the evidence beneath it.

Ledger

The Changes lens at its landscape depth: every change with its current tier.

Lens

One view of your data with its own landscape: Prompts, Competitors, Changes, Sources, Intelligence, Trace, Industry Pulse, Readiness.

Negative mention

An answer that names you in order to warn readers off or steer them elsewhere. Recorded with negative sentiment and not counted as a citation.

Not measurable

A change that cannot be scored, with a named reason. Not a zero and not a failure.

Observed live

A change tier: our own fetch confirmed the page satisfies the change.

Parent organisation

An agency's account. Clients live inside it as workspaces.

Prompt

A question you want to be the answer to, phrased the way a buyer would ask it. Never contains your own name, except on deliberately branded prompts.

Proof

The lens showing the verbatim sentences engines wrote about you, with surface and date.

Readiness

The grade on your own site across four levels, expressed as a tier: Invisible, Emerging, Cited, Preferred.

Share of voice

Your portion of all business naming across your tracked prompts.

Source

A page or domain an engine leaned on when writing an answer. Counted, not inferred.

Surface

Any of the seven places measured. Six are chat engines; the seventh is AI Overviews.

Topic

A group of prompts about one buyer problem. How the Prompts lens navigates and how reporting rolls up.

Trace

The lens for arrivals from AI answers, and for AI crawler fetches when the edge worker is installed.

Verifier

The independent model that audits the judge whenever the judge is not certain. Disagreements are not counted.

Workspace

One brand or one client: its profile, prompts, competitors, citations, changes and reports, isolated from every other.

Zero-click

A search that ends without a visit, because the answer was given in place. The reason citations matter more than referral traffic.

Words we avoid

Keywords, because a prompt is a question, not a phrase. Rankings and SERPs, because an AI answer is not a list of positions. Industry average, because we publish no benchmark we cannot source.

Related