Reference
Market Authority Score (MAS)
A governed, multi-signal composite reading of observed authority on AI/search surfaces — built from presence, position, citation, source, authority, and surface signals.
This page defines the Market Authority Score as a public reference concept. It is descriptive and does not disclose proprietary scoring mechanics.
Canonical Reference · Reference Library · v1.2 · Last reviewed 2026-06-03
Applies to: Observatory readings that present or interpret a Market Authority Score
Status: Definitions govern downstream Observatory usage
Quick answer
The Market Authority Score is a governed composite reading of an actor's observed authority on AI and search surfaces, within a defined scope and time window. The Market Authority Score is built from multiple observed signal families — not from a single number, and not from a count of mentions. The score reads how an actor is treated on the surface; it is not a measure of the actor's business market.
Core definition
The Market Authority Score is a governed composite reading of an actor's observed authority on AI and search surfaces, within a defined scope and time window — built from multiple observed signal families, not a single number or a count of mentions.
Interpretation rules
- The score is a composite of observed signal families (presence, position, citation, source, authority, surface), read together.
- It reads how an actor is treated on the surface, never the business market behind it.
- The signal families stay separately inspectable; the narrative layer (Deep Narrative Capture) is never folded into the score.
- The public layer describes the signal families and selected readings; exact weights, formulas, thresholds, and the full query universe are not published.
Purpose & scope
Why this page exists. Counting mentions answers "how often"; the score answers "with what authority, on which surface" — a reading that needs several signals read together rather than one figure.
Where it applies. Every reading that presents or cites a Market Authority Score applies this definition.
What the score is built from. The Market Authority Score combines several observed signal families: presence, position, citation behavior, source behavior, authority signals, and surface readings. Each family answers a different question, and the composite reads them together — because, on AI surfaces, position is more meaningful when read alongside citation and framing.
What the score is not. The Market Authority Score is not a visibility count: counting mentions answers "how often"; the score reads "with what authority, on which surface." The score is not a simple visibility metric or a single unqualified number, and it is not a measure of revenue, market share, or commercial leadership. It is a surface-authority reading, scoped and dated.
How the score stays governed. The signal families behind the Market Authority Score remain separately inspectable rather than collapsed into one opaque figure, the method is held stable across comparable windows, and every published score carries an attestation. The narrative layer (Deep Narrative Capture) is never folded into the score.
What is published, and what is not. The public layer describes the signal families and selected composite readings. The exact weights, formulas, thresholds, and the full query universe behind the Market Authority Score are not published. This balance is deliberate: public enough to prove the reading is serious and governed, incomplete enough to preserve the intelligence layer.
What it is / what it is not
It is
- a governed, multi-signal composite reading of surface authority, scoped and dated
- built from separately inspectable signal families
It is not
- a visibility count or a single unqualified number
- the narrative layer (which is never folded in)
- a measure of revenue, market share, or commercial leadership
Common misinterpretations this page prevents
“The score is a mention count or a visibility metric.”
Counting mentions answers "how often"; the score reads "with what authority, on which surface." It is a composite surface-authority reading, scoped and dated, not a simple visibility tally.
“A high score means the actor leads its business market.”
The score reads how an actor is treated on AI and search surfaces within a scope and window. It is not a measure of revenue, market share, or commercial leadership.
“Publishing a score means publishing the method.”
The public layer describes the signal families and selected readings; the exact weights, formulas, thresholds, and full query universe are not published — public enough to prove the reading is governed, incomplete enough to preserve the intelligence layer.
How this page governs downstream readings
Where a downstream output treats the score as a raw count or a commercial verdict, this definition prevails:
- An Authority verdict that cites a score carries its scope, window, and limitation.
- A card that names a score does not restate it as market leadership.
- A Decision Guide reads the score as a composite surface reading, not a mention count.
Relationship to other references
- AI Behavior Intelligence — the score is the governed composite of the signals behavior intelligence reads.
- Deep Narrative Capture — the narrative overlay is kept out of the score by design.
- Official vs Narrative Layer — the score lives in the official measured layer.
- Attestation Standard — every published score carries an attestation.
- Benchmark Governance — method stability is what keeps the score governed.
Canonical notice
This is a canonical reference. It defines how the Market Authority Score is read and published in Observatory public readings, and it does not disclose the score's proprietary weights, formulas, thresholds, or query universe. That definition governs every downstream use across the site. It does not itself rank, promote, or measure commercial outcomes. Where an editorial page, a card, or a market reading appears to conflict with this definition, this reference prevails.
Intelligence by AAI Kernel. Published by The Observatory.
The Market Authority Score is an observed reading on defined AI and search surfaces within a stated scope and time window. The score is not a measure of real-world revenue, market share, or business leadership.
Related references
- AI Behavior Intelligence — because the score is the governed composite of the signals it reads.
- Deep Narrative Capture (DNC) — because the narrative overlay is never folded into the score.
- Official vs Narrative Layer — because the score lives in the official measured layer.
- Attestation Standard — because every published score carries an attestation.
- Benchmark Governance — because method stability is what keeps the score governed.