Definition

What is Citation Trust?

Not all AI citations are equal. Citation Trust is SearchScore's measure of how durable yours are: the difference between durable visibility and a lucky mention.

Citation Trust is a measure of AI citation quality, coined by SearchScore. It reports what share of your AI citations come from engines reading your live pages versus their training memory, how many of the six tracked engines cite you at all, and how often you land in a ranked list versus loose prose.

A raw citation count answers one question: did AI mention you? Citation Trust answers the harder one: can you rely on it happening again? Two businesses with the same number of citations can be in completely different positions, and Citation Trust is how the SearchScore Tracker tells them apart.

The three things it measures

Source provenance What share of your citations come from AI reading your live pages, versus repeating whatever its training memory holds about you. Live-page citations respond to the site you control today; memory citations reflect a snapshot you cannot edit.
Engine breadth How many of the six tracked engines (ChatGPT, Claude, Gemini, DeepSeek, Grok and Perplexity) cite you at all. Visibility concentrated in one engine is exposure to one provider's next update.
Structured-citation rate How often you appear in a ranked list of recommendations versus being mentioned in loose prose. A ranked placement is a recommendation; a passing mention is not.

Why provenance is the part that matters most

AI engines surface a brand from two different places: by reading its live pages at answer time, or by recalling it from training memory. The split matters because only one of them is under your control. When your citations come mostly from live pages, the work you do on your site feeds directly into future answers. When they come mostly from memory, you are trading on a frozen snapshot: it can carry outdated facts, and it can fade when the model is next updated, without anything on your site changing.

This is also why the Tracker's AI Visibility Score weights each citation by whether AI cited your live site or its training memory, alongside sentiment and position. A high score built on live-page, ranked-list citations across several engines is defensible; the same number built on memory mentions in a single engine is fragile, and Citation Trust is the view that exposes the difference.

Where you see it

Citation Trust appears as a dedicated strip on the Tracker dashboard, showing your live-versus-memory split, your engine breadth out of six, and your structured-citation rate, updated with every scan (weekly on Starter, daily on Pro). If you want to see how it fits into the full monitoring picture for one engine, start with the ChatGPT visibility tracker overview.

How to read the three numbers together

Individually, each component tells a partial story; together they describe how much weight your AI visibility can bear. A brand cited by five of six engines, mostly from live pages, mostly in ranked lists, has visibility that survives a model update, because the engines keep re-reading a site that keeps earning its place. A brand cited by one engine, from memory, in passing prose has a statistic, not an asset: nothing it does this quarter will change that mention, and the next model release could erase it. When Citation Trust is weak, the levers are the same ones the Tracker's Action Centre ranks for you: make your live pages readable and quotable enough that engines cite them directly, on more engines, in the list placements buyers actually act on.

See your own Citation Trust.

The Tracker scans six AI engines and shows where your citations really come from. From £20/mo, cancel anytime.

Explore the Tracker →