AEO Glossary

Overview

This glossary is a comprehensive reference for Answer Engine Optimization terminology, metrics, and intelligence scores. Whether you are new to AI visibility monitoring, evaluating AEO tools, or an experienced practitioner, these definitions cover every concept you will encounter when tracking and optimizing how AI models represent your brand.

Terms are organized alphabetically. Each definition explains what the term means, why it matters, and how it is used in practice.

---

A

AEO (Answer Engine Optimization)

The practice of optimizing your brand, content, and online presence to appear accurately and prominently in AI-generated answers. AEO is to AI chatbots (ChatGPT, Claude, Gemini, Perplexity, Grok) what SEO is to search engines (Google, Bing). While SEO focuses on ranking in search results, AEO focuses on being cited, mentioned, and recommended when AI models respond to user queries. See What is AEO? for a full introduction.

AEO Score

A rating from 0 to 100 that measures how well a specific web page is optimized for AI citation. The score evaluates schema markup, entity clarity, content structure, FAQ presence, and other factors that influence whether AI models will reference the page when generating answers. Calculated by the Page Optimization tool within AEO Optima.

Action Verification

The process of measuring whether a specific optimization action actually improved AI visibility. AEO Optima tracks each recommended action through its lifecycle — recommendation, implementation, and impact measurement — then correlates the action with visibility changes to determine its efficacy. Over time, this creates an evidence base of what works for your brand and category.

AI Analysis

AI-powered deep analysis of snapshot patterns that goes beyond basic mention detection. Includes sentiment drivers, content gaps, opportunity scoring, and comprehensive analysis modes. Results generate actionable insights with specific recommendations for improving brand visibility across AI platforms.

AI Brand Score

A composite metric that represents your brand's overall health across AI answer engines. It is a weighted blend of five signals — Visibility (35%), Sentiment (15%), BNCI / brand-narrative coherence (15%), CMCS / cross-model consistency (15%), and ETAS / topical authority (20%) — renormalized over whichever signals have enough data. The AI Brand Score provides a single number that tracks your brand's AI presence over time, and is the headline number on generated reports.

AI Visibility

The degree to which your brand appears in AI-generated answers across different models and query types. AI visibility is distinct from traditional search visibility — a brand can rank #1 on Google and still have zero AI visibility if AI models do not mention it in their responses. AEO Optima measures AI visibility as a percentage of monitored prompts where your brand is mentioned.

Anomaly Detection

Completeness-aware statistical detection of unusual changes in visibility, sentiment, or mention-rate metrics. Uses robust (median/MAD) Z-score analysis with Benjamini-Hochberg false-discovery-rate control across the three simultaneous tests. Only analyzes days with sufficient data (at least 80% of expected snapshots and at least 10 snapshots absolute), excludes the current day, and marks anomalies as persistent when two consecutive points are anomalous. This prevents false alarms from sparse data or single spikes.

B

BNCI (Brand Narrative Coherence Index)

One of AEO Optima's 6 proprietary intelligence scores. BNCI measures how consistently AI engines tell your brand's story across different prompts and models. A high BNCI means AI engines have a unified understanding of your brand narrative — they describe your value proposition, differentiators, and positioning in a consistent way. A low BNCI indicates fragmented or contradictory narratives that may confuse users who ask different AI models about your brand.

Bootstrap Prediction Interval

A 95% prediction interval for visibility forecasts built by resampling the forecasting model's residuals 500 times and taking the 2.5th and 97.5th percentiles. This is an honest alternative to assumed-normal intervals — the interval widens naturally where the model is uncertain, without making distributional assumptions that may not hold for AI visibility data.

Brand Facts

Verified brand attributes (founding year, headquarters, key products, pricing, market position, etc.) stored in your project settings. Brand facts serve two purposes: the accuracy checker compares AI responses against them to detect hallucinations, and the entity analyzer uses them to score how clearly AI models understand your brand's identity.

Brand Overlap

A metric that measures how often your brand and a specific competitor appear together in the same AI-generated response. High brand overlap indicates that AI models frequently consider both brands in the same competitive context — useful for understanding which competitors AI models associate with your brand.

Branded Prompt

A prompt segment classification for queries that mention your brand name but no competitors (e.g., "What do people think of TechShu?" or "Is AEO Optima good for agencies?"). Analytics filtered to branded prompts show how AI engines describe your brand when users explicitly ask about you.

Building Block

A reusable query template component in the Query Universe system. Building blocks are categorized by type — Core Services, Modifiers, Audiences, Intents, Geographies, and more — and combined by the prompt composer to systematically generate monitoring prompts. This ensures your monitoring covers every dimension of your brand's query space without manual prompt creation.

C

Capture Run

The receipt for one batch of snapshot captures. Every manual "Run now" and every scheduled capture records the exact prompt × model × location shape it executed, expected vs actual snapshot counts, duration, status, and the run's surface mix (chat LLMs vs Google AI Overview vs AI Mode vs organic SERP). Shown in the product as a cycle: the Cycles tab of Run schedule & cost lists them, and two cycles that covered exactly the same questions, engines and locations can be diffed side-by-side, so metric moves are always compared like with like. See Cycles.

CIPS (Citation Impact & Positioning Score)

One of AEO Optima's 6 proprietary intelligence scores. CIPS measures your brand's citation footprint as a fixed three-part composite: a citation-rate component (0–40, an exponential-saturation curve on how often AI answers cite your brand), a hub-influence component (0–30, your brand's share of total citation-hub influence), and a source-diversity component (0–30, the Shannon-entropy spread of the source types citing you, weighted by sample confidence). A high CIPS means your brand is cited often, from influential hubs, across a healthy diversity of sources.

Citation Gap

A topic or query area where competitor brands receive citations from AI engines but your brand does not. Citation gap analysis identifies specific sources, publications, or content types that competitors leverage for AI visibility — giving you actionable outreach targets to close the gap. See also CIPS.

CMCS (Cross-Model Consistency Score)

One of AEO Optima's 6 proprietary intelligence scores. CMCS measures how consistently your brand is represented across different AI models (ChatGPT, Claude, Gemini, Perplexity, Grok, and others). A high CMCS means all AI models describe your brand similarly. A low CMCS indicates that some models have incomplete, outdated, or conflicting information — a common situation that requires model-specific optimization strategies.

Competitor Prompt Segment

A prompt segment classification for queries that mention both your brand and one or more competitors, or contain a "brand vs X" comparison pattern (e.g., "AEO Optima vs other AEO tools"). Analytics filtered to competitor prompts reveal your head-to-head positioning in AI-generated comparisons.

Completeness-Aware Detection

A safeguard in AEO Optima's anomaly detection system that excludes days where snapshot capture was partial. A day is analyzed only when it has both at least 80% of expected snapshots — where "expected" is the observed median of positive daily capture counts, self-calibrating to the project's real cadence — and at least 10 snapshots absolute. This prevents the detector from flagging apparent "visibility drops" that are actually just incomplete data collection.

Confidence Quality Rating

A High, Medium, or Low signal attached to every visibility forecast. Derived from cross-validated prediction error, coverage probability of prediction intervals, and the Ljung-Box residual autocorrelation test. The confidence quality rating tells you how much weight to place on the forecast's point estimate and interval bounds.

Connector

A third-party integration that extends AEO Optima's data collection and automation capabilities. Supported connectors include Serper (SERP data), DataForSEO (search analytics), Bing, Brave Search, Google Knowledge Graph, Wikipedia, Reddit, Hacker News, Stack Exchange, GitHub, YouTube, GDELT, WordPress, Shopify (e-commerce data), Slack (notifications), Looker (BI dashboards), and Zapier (workflow automation).

Coverage Report

An analysis generated by the Query Universe system showing how thoroughly your building blocks and prompts cover your brand's query space. Coverage reports include dimension distributions, gap identification, and specific recommendations for expanding monitoring coverage into underrepresented areas.

D

Domain Authority (AI Citation)

A measure of how frequently a domain is cited by AI engines as a source. Calculated from citation frequency, recency, and cross-model presence. Higher domain authority means AI models trust and reference that source more often — making it a valuable signal for identifying which sources to target in your citation-building strategy.

E

Entity Clarity

A score from 0 to 100 that measures how clearly AI engines understand your brand's identity and attributes. It blends four signals: attribute richness, knowledge-graph completeness, cross-model consistency, and the depth of relationships AI associates with your brand (minus a penalty for entity confusions). The knowledge-graph completeness signal checks each fact you record in brand facts — whatever its label — against the AI responses that mention you, and counts the fact as "known" when the AI's answers corroborate its value. This works for any fact you enter (founders, awards, services, locations, parent company, and so on), not a fixed list, so completeness reflects the real share of your facts the AI demonstrably knows. Higher entity clarity means AI engines have accurate, detailed, and consistent knowledge of who your brand is and what it offers.

ETAS (Entity & Topical Authority Score)

One of AEO Optima's 6 proprietary intelligence scores. ETAS measures how strongly AI engines associate your brand with specific topics, products, or expertise areas. A high ETAS indicates that AI models recognize your brand as an authority in its domain — they recommend you for relevant queries and accurately describe your areas of expertise.

F

Fusion Insights

Insights produced by joining several first-party and AI-visibility signals on your own page URLs: AI snapshot citations, GA4 AI-referral traffic, Google Search Console rank and first-party AI-appearance (AI Overviews / AI Mode), optional DataForSEO AI-Overview share, and social/community mentions. Fusion closes the loop from "an AI cited this page" to the real traffic, rank, and demand that citation drives — surfacing insights like cited-page-driving-traffic, cited-page-at-risk, AI-vs-SEO divergence, citation-gap-with-demand, and AI-Overview-vs-organic gaps. It is plug-and-play: only the sources you've connected contribute, and each insight degrades gracefully when a source is missing. Surfaced on Analytics, Citations, Revenue, the Insights → Discover feed, generated reports (the "AI Visibility ↔ Real Traffic" section), and the get_fusion_insights MCP tool. See Fusion Insights.

G

GEO (Generative Engine Optimization)

The practice of optimizing web content specifically for generative AI engines to cite and reference. While AEO is the broader discipline of brand visibility in AI answers, GEO focuses on the technical and content-level factors that make individual pages more likely to be cited. AEO Optima's GEO Audit analyzes pages across schema markup, entity clarity, FAQ structure, content depth, technical SEO, and freshness to produce a readiness score with actionable recommendations.

GEO Audit

A multi-dimensional assessment of a web page's readiness for AI citation. The audit evaluates schema markup, entity clarity, FAQ structure, content depth, technical SEO signals, and content freshness. Each dimension receives a score, and the combined result indicates how likely AI models are to cite the page. Includes specific, actionable recommendations for improving each dimension.

H

Hallucination

An AI-generated statement about your brand that is factually incorrect. Detected by comparing AI responses against your verified brand facts. Examples include wrong founding dates, incorrect product descriptions, or fabricated statistics. Hallucinations can be flagged via the correction submission workflow to notify AI providers of inaccuracies in their models.

Holt-Winters Ensemble

The statistical forecasting engine used by AEO Optima's visibility forecaster. Trains three competing models — Holt-Winters additive (level + trend + weekly seasonality), Holt damped trend, and Seasonal Naive — grid-searches parameter combinations optimized by AICc, then ensembles them weighted by expanding-window time-series cross-validation RMSE. Produces point forecasts plus bootstrap-calibrated 95% prediction intervals and a confidence quality rating.

I

Intelligence Engine

One of AEO Optima's 5 daily computation engines that process captured snapshot data into actionable intelligence. The five engines are: citation impact analysis (tracing source influence), competitor trajectory tracking (monitoring competitive movement), prompt decomposition (understanding query patterns), error root-cause detection (identifying why visibility drops), and leading indicator identification (predicting future changes before they appear in headline metrics).

Intelligence Scores

AEO Optima's 6 proprietary brand health metrics that measure distinct dimensions of AI visibility. The six scores are: BNCI (narrative coherence), CMCS (cross-model consistency), MEI (market volatility), SDI (sentiment stability), CIPS (citation impact), and ETAS (topical authority). No other AEO tool provides these metrics. See the AEO Tools Comparison for context.

L

Leading Indicator

A metric or signal that predicts future changes in AI visibility before they appear in headline metrics like mention rate or visibility score. AEO Optima's leading indicator engine identifies early warning signals — such as shifts in citation patterns, changes in competitor mention frequency, or sentiment trend inflections — so you can act proactively rather than reactively.

M

MCP (Model Context Protocol)

An open standard for connecting AI assistants to external tools and data sources. AEO Optima provides 119 MCP tools that allow AI clients — including Claude Desktop, Claude Code, ChatGPT, Cursor, VS Code Copilot, Windsurf, Gemini, and Amazon Q — to directly query visibility data, run analytics, capture snapshots, and generate reports. See the MCP API reference for details.

MEI (Market Entropy Index)

One of AEO Optima's 6 proprietary intelligence scores. MEI measures how fragmented vs. concentrated the competitive landscape is within AI answers — the Shannon entropy (with an HHI companion) of the brand-mention mix at a point in time. High entropy means many brands share the space in AI answers (a fragmented, open market); low entropy means a few brands dominate (entrenched positions). It is a cross-sectional concentration measure, not a measure of how recommendations change over time — that temporal view is SDI's domain.

Mention Rate

The percentage of monitored prompts where at least one AI engine mentions your brand in its response. Distinct from visibility score in that mention rate can be calculated per-model, per-segment, or per-time-period, while visibility score is the primary aggregate metric shown on the dashboard.

Milestone

A defined checkpoint within a goal-based optimization plan. Milestones represent intermediate targets on the path to a visibility goal (e.g., "reach 40% mention rate" as a milestone toward a 60% goal). Each milestone includes a target metric value and a deadline, with pace indicators showing whether you are on track, ahead, or behind.

N

Non-Branded Prompt

A prompt segment classification for queries that mention neither your brand nor any competitor (e.g., "Best project management tool for remote teams" or "Top CRM for small business"). Non-branded prompts are the most important segment for measuring true organic AI discoverability — they reveal how often AI recommends your brand when users are not specifically asking about you.

P

Pace Status

An indicator within goal tracking that shows whether progress toward a visibility milestone is on track, ahead of schedule, or behind schedule. Calculated by comparing actual metric improvement against the expected trajectory based on milestone dates and target values. Pace status helps teams prioritize optimization efforts on goals that need attention.

Persistent Anomaly

An anomaly where both the flagged data point and the immediately preceding point exceed the z-score threshold. A single spike is often noise caused by data variability; two consecutive anomalous points indicate a genuine trend change. Use the persistent flag to filter out one-off events in alert rules and focus on meaningful shifts.

Prompt Segment

One of three classifications automatically assigned to every monitoring prompt: Branded, Non-Branded, or Competitor. Auto-detected from brand name, competitor names, and "vs" comparison heuristics, with manual override available. All analytics surfaces in AEO Optima accept a segment filter so you can view any metric for any slice of your prompt portfolio.

Q

Query Universe

A systematic framework for generating, organizing, and managing all possible brand-related queries that AI models might encounter. The Query Universe uses building blocks — reusable template components categorized by type — combined by a prompt composer to ensure comprehensive monitoring coverage across topics, intents, audiences, geographies, and modifiers. Coverage reports identify gaps in your monitoring. See Core Concepts for a guided introduction.

R

Rank Position

The position where your brand appears in a numbered or ordered list within an AI-generated response. A rank of 1 means your brand was mentioned first — the strongest position. Not all AI responses contain ranked lists; this metric only applies when the AI model organizes its answer as an ordered recommendation. Tracking rank position over time reveals whether your brand is moving up or down in AI-generated rankings.

S

Selection criteria

A per-engine map of what each AI engine believes defines the best offering in your category and industry — derived by asking the engines themselves and structuring the answers into three criterion kinds: gates (pass/fail, applied first — fail one and the candidate is dropped), segment criteria (vary by buyer and context), and signals (lift but never eliminate). Each gate is matched against your approved brand facts to produce the gate-clearance table that leads the page. The map is a stated worldview at derivation time, not a measured frequency. See Selection criteria.

SDI (Sentiment Drift Index)

One of AEO Optima's 6 proprietary intelligence scores. SDI tracks how AI sentiment toward your brand changes over time. A stable SDI indicates consistent positive or neutral perception across AI platforms. A volatile SDI signals shifting AI opinion — possibly driven by new training data, recent news, or competitor activity — that may require attention before it becomes a reputation problem.

Sentiment

The overall tone an AI model uses when describing your brand in its response. Classified as Positive (favorable language, recommendations, endorsements), Neutral (factual mention without strong opinion), or Negative (critical language, warnings, unfavorable comparisons). Sentiment is tracked per-snapshot and aggregated over time to reveal trends in how AI models perceive your brand.

Share of Model

The platform's core visibility measurement under its AEO/GEO-framework name: the percentage of sampled AI runs (per prompt and engine) in which your brand appears. It is an inclusion frequency, never a rank — AI answers form a stochastic distribution, so frequency across many runs is the honest unit of measurement. Readings carry their run counts, with explicit flags for thin samples (low confidence under 10 runs, below target at 10–29; the discipline targets 30–50 runs per prompt and engine), and engines are reported per engine rather than averaged. Same computation as mention rate. See Analytics → Share of Model.

Mention share

Also known as share of voice. Your brand's proportion of total brand mentions across all AI responses in your monitoring set. If AI engines mention 5 brands across your prompts and your brand appears in 40% of those mentions, your mention share is 40%. This competitive metric reveals your relative AI presence compared to competitors within your category.

Snapshot

A single point-in-time capture of an AI model's response to one of your monitoring prompts. Each snapshot records the full response text, whether your brand was mentioned, sentiment classification, rank position, citation sources, token usage, and cost. Snapshots are the fundamental data unit in AEO Optima — all analytics, intelligence scores, and forecasts are derived from snapshot data.

V

Visibility Forecasting

Statistically rigorous prediction of future AI visibility trends using the Holt-Winters Ensemble engine. Forecasts include point estimates, bootstrap-calibrated 95% prediction intervals, confidence quality ratings, and diagnostic information. Requires a minimum of 14 CAPTURES — 14 days for a project capturing daily, 14 weeks for one capturing weekly. Seasonal models need three further weeks and are offered only to daily-capture projects. Where there are captures but no recent ones, no forecast is returned at all and the reason is given in words. Forecasting helps teams set realistic goals and anticipate visibility changes before they occur.

Visibility Score

A composite 0–100 score that blends four signals into a single quality measure of overall AI brand presence — mention rate (~44%), rank-position quality (~28%), sentiment (~17%), and model-coverage breadth (~11%). This is the primary headline metric on the AEO Optima dashboard. It is distinct from mention rate, which is simply the raw percentage of responses that mention your brand; the Visibility Score weights that frequency together with where and how favorably you are mentioned, so a brand mentioned often but ranked low or described negatively scores lower than its raw mention rate alone would suggest. Soft-deleted snapshots (for example, those removed when a prompt is deleted) are excluded from the computation on every path, so the score reflects only active data.

---

Related Resources

• What is AEO? — Understanding answer engine optimization and why it matters
• Core Concepts — A guided introduction to the foundational ideas behind AEO Optima
• AEO Tools Comparison — How AEO Optima compares to other AI visibility approaches
• Supported AI Models — Details on the AI providers and models available for monitoring
• Best Practices — Strategies for improving your AI visibility metrics