Why a public metric standard matters
AI visibility becomes meaningless if every vendor defines it differently. SIGNAL publishes definitions so brands, agencies and auditors can compare results consistently and see where direct observation ends and modeling begins.
AI Visibility
Definition: scoped measure of how consistently an entity appears across relevant AI answer opportunities.
Answers: “Does the brand show up where it should?” It is not equivalent to website traffic, search ranking or ad impressions.
Mention Rate
Share of relevant measured answers where the entity is named. A mention can be positive, negative or neutral and does not automatically imply recommendation.
Recommendation Rate
Share of relevant measured answers where the entity is recommended or selected as an appropriate option according to the methodology. A recommendation is stronger than a mention.
Answer Position
Relative prominence or order of an entity within a ranked, comparative or recommendation answer.
Citation Rate
Frequency with which a defined source, domain or entity is cited in measured answers.
Citation Coverage
Share of relevant answer opportunities where eligible sources appear as citations.
Source Influence
A measure or model describing the relationship between information sources and AI answers. Source influence may be modeled; direct citation is not automatically causal influence.
Topic, intent and narrative coverage
Topic Coverage
Share of strategic topics where the entity appears with sufficient relevance.
Intent Coverage
Share of defined intent families — such as learn, compare, buy, switch, troubleshoot, local recommendation or trust/reputation — where the entity appears.
Narrative Frequency
How often a recurring claim or description appears across measured answers.
AI Sentiment / Characterization
How the entity is described along defined dimensions. Prefer dimension scores or classifications over a single simplistic sentiment number.
Estimated AI Demand
Estimated quantity or relative level of demand for a prompt/intent family. It must be labeled estimated unless provided directly by an authorized AI platform.
Prompt Trend
Change in demand or observed frequency for a prompt/intent family over time.
AI Crawler Requests
First-party observed server requests classified as originating from known or likely AI crawlers/agents.
Content Retrievals
Successful machine requests that retrieve a defined page, file, content object or endpoint.
AI Agents Detected
Distinct identified AI crawler/agent identities observed within scope. Identity methodology is documented in Methodology.
AI Referral Sessions
Human sessions attributed to a known AI referral source using first-party or authorized evidence.
AI Referral Conversion Rate
Conversions associated with measured AI referral sessions divided by those sessions under the defined attribution rule.
AI Referral Revenue
Revenue attributed to AI referral sessions under the selected attribution model.
Crawler-to-Referral Ratio
A descriptive comparison of machine access volume to AI-driven human referral volume. It does not prove that crawling caused referrals.
Claimed Uplift
Improvement claimed by a vendor, agency, platform or client before independent verification.
Measured Uplift
Change measured by SIGNAL under a named methodology.
Incremental Lift
Estimated causal effect relative to an appropriate counterfactual or control. Use only when methodology supports a causal conclusion.
Claim Support Status
Verify classifications: Supported, Partially supported, Unsupported, Inconclusive or Not measurable.
Evidence Completeness
Extent to which required data and evidence are present for an audit.
Policy Match
Whether observed crawler/agent access behavior matches the intended access policy.
AI Media Efficiency and candidate indices
AI Media Efficiency is a working concept comparing AI-related media spend to defined business outcomes. Its exact public formula must be approved before use as a standardized index.
Candidate product indices such as AVI, ASOV, ARR, CAS, AAS, AME and IAR should be promoted as standardized indices only after formulas are validated and versioned.
Every metric card needs a scope label.
AI Share of Voice: 31% Universe: 1,200 sampled answer opportunities Platforms: selected supported AI systems Markets: US, UK Language: English Competitors: Brand A, B, C, D Period: Aug 1–31, 2026 Methodology: v1.3 Evidence: controlled external measurement
The values above are an illustrative scope example, not a market benchmark or customer result.
Use the definition before using the number.
Pair every SIGNAL result with the relevant scope, evidence class and methodology.