What is the AI Presence Score?

The AI Presence Score is a composite summary of how the benchmark represented the company. It combines several structured signals from the captured answers, including coverage, competitive outcomes, claim accuracy, sentiment, source evidence, and confidence markers.

It is written at scoring time from the same records used by the report drill-downs. The score is not a separate model opinion layered on top.

What should you read under the score?

Start with the captured answers. Then inspect material claims, competitive outcomes, misrepresentations, available sources, and the journeys where the pattern appears. That decomposition explains which part of the representation needs attention.

A topline score can move while one important journey gets worse. The journey record is the useful operating grain for the managed cycle.

Is a later score change proof of the content result?

No. The default statement is changed, then measured movement. Other pages, competitors, model updates, retrieval changes, and normal answer variation may also be present.

A qualifying holdout can permit a narrower causal statement for the contrast it supports. Headline workspace and journey rates remain association-grade.

What happens when the sample is thin?

KnitKnot applies a reporting floor before printing a delta. A journey or cycle below that floor is labeled insufficient data. The state is shown even when a directional result would make a better story.

How does the score fit Get Agent Ready?

The score belongs to the benchmark. KnitKnot preserves that before-state, manages evidence-backed content on eligible pages, and connects the later measurement window to the published version and cycle.

Read the full measurement methodology.