Metrics & Scores

PIS (Prompt Impression Score) is the primary AI visibility metric, measuring what percentage of relevant user prompts result in your brand being mentioned by AI.

Also: Prompt Impression Score

SMR (Semantic Match Rate) measures the alignment between your content and the underlying intent of user queries.

Also: Semantic Match Rate

ARD (AI Reasoning Depth) measures how thoroughly AI explains and justifies its recommendation of your brand.

Also: AI Reasoning Depth

PPMA (Prompt-Page Mapping Accuracy) measures how comprehensively your content addresses the questions users are asking.

Also: Prompt-Page Mapping Accuracy

Context Resilience Score (CRS) measures how stable your AI visibility is when prompts are modified with different contexts, constraints, or noise.

Also: CRS, Stability Score

Model consensus measures whether different AI models provide consistent information about your brand.

Also: Cross-model consistency

AI sentiment measures the tone and framing AI uses when discussing your brand.

Also: AI tone, Brand perception

Cross-model agreement measures how often different AI models recommend the same brands for the same query.

Also: Platform agreement, AI consensus

Temporal instability refers to variance in AI recommendations over time.

Also: AI response variance, Recommendation volatility

PASOR (Platform-Adjusted Stability of Recommendations) is a research metric that measures how consistently an AI platform recommends brands across multiple queries.

Also: Platform stability score

Citation share measures how often AI platforms explicitly reference and link to your content when generating answers.

Also: Source share, Citation rate

Answer share measures your brand's presence in AI-generated answer text, regardless of whether you are cited as a source.

Also: Response share, Mention share

Visibility velocity measures the change in your AI visibility metrics over time.

Also: Visibility trend, Growth rate

Recommendation momentum tracks whether AI is recommending your brand more or less frequently over time.

Also: AI preference trend

Persona-AI alignment measures whether AI recommends your brand to the right audiences.

Also: Audience alignment, ICP-AI fit

Brand generosity score measures how many brands an AI platform typically includes when answering recommendation queries.

Also: Platform brand density, Mention capacity

An AI impression occurs each time your brand appears in an AI-generated response to a user query.

Retrieval rate measures the percentage of times your content is retrieved by AI systems using Retrieval-Augmented Generation (RAG) when users ask questions relevant to your brand or category.

Query fan-out refers to the background queries that AI systems generate internally to gather comprehensive information for a single user question.

Narrative control score measures the degree to which a brand influences what AI says about it -- comparing the AI's actual narrative about your brand against your intended brand messaging.

Persona resonance score measures how effectively AI connects your brand to specific buyer personas when those personas ask questions relevant to your products or services.

The AI Reputation Index (ARI) is a 0-100 score of how AI models describe and recommend a brand, combining sentiment, recommendation share, source quality, consistency, and stability.

Also: ARI, AI reputation score

The bilingual penalty is the gap between how AI rates a brand in its home language versus in English, ranging from -7.6 points (penalized at home) to +23.1 points (boosted at home) in the Rankfor Index 2026.

Also: Home-language AI gap, Cross-language ARI delta

The frequency at which AI mentions your brand when answering relevant queries.

Also: Mention frequency, AI recall rate

The Category Ownership Index (COI) measures how thoroughly one brand dominates the buyer-intent questions of its category in AI recommendations, from 0 (absent) to 1 (owns every question).

Also: COI, category ownership

Measures how reliably AI gives the same answer when asked the same question multiple times.

Also: Response consistency, Answer reliability

Long-tail citations are the wide set of low-frequency domains that AI uses to ground answers, following a Zipf distribution where the n-th-ranked source receives roughly 1/n^0.85 of citations.

Also: Citation long tail, Zipf citation distribution, AI link economy long tail

Measures whether AI uses the same talking points and narrative structure across multiple responses.

Also: Content similarity, Narrative consistency