AI Visibility & Machine Relations Glossary

An architectural pattern where an LLM-powered agent dynamically controls the information retrieval process, deciding when to search, what to query, which tools to use, and whether the results are sufficient before generating an answer. Agentic retrieval replaces the static one-pass pipeline of traditional RAG with an iterative loop that can decompose complex queries, execute parallel sub-searches, self-evaluate results, and re-retrieve until the evidence meets a quality threshold.

The aggregate trust signal a brand accumulates across AI engines that determines whether it gets recommended — not just recognized — in AI-generated answers. AI brand authority is the compound of third-party evidence, entity consistency, and cross-platform citation density that separates brands AI engines actively suggest from brands they merely identify when asked.

AI Visibility is how often ChatGPT, Perplexity, Gemini, and Google AI cite your brand in buyer answers. Only 11% of domains get cited across multiple AI engines. Definition, measurement framework, and the five levers that actually move it.

The discipline of structuring content and brand signals so that AI-powered answer engines — Perplexity, Google AI Overviews, ChatGPT, Gemini, Claude — select your brand as the cited source in direct answers to user queries.

The emerging attention system where AI engines distribute discovery and demand through citations rather than links — characterized by winner-take-most dynamics and radical concentration.

Citation eligibility is the threshold a source must clear before AI retrieval systems will select it as a cited reference in generated answers. Where retrieval eligibility determines whether AI can find your content, citation eligibility determines whether it will use your content as evidence. The distinction matters: a page can be crawlable, indexed, and structurally extractable, yet never appear as a cited source because it fails the trust, corroboration, or answer-quality gates that AI engines apply after retrieval.

A strategic communications discipline focused on earning online media coverage, authoritative backlinks, and brand mentions that build domain authority, search visibility, and increasingly, AI citation eligibility. Digital PR replaces broadcast-era impressions with measurable digital signals: referring domains, entity mentions, and the structured evidence that AI retrieval systems now evaluate when deciding which brands to cite.

Third-party credibility signals — tier-1 media placements, expert citations, independent reviews — that AI engines weight far more heavily than brand-owned content when deciding which brands to recommend.

Media coverage, mentions, and citations that a brand receives through editorial merit rather than paid placement — the primary source material that AI engines use when deciding which brands to recommend.

An entity chain is the network of independent, cross-domain mentions that lets AI retrieval systems verify that a brand, concept, or claim exists across multiple trusted contexts before citing it. Unlike backlinks, which pass authority through a single hyperlink, entity chains create a distributed verification layer where each independent mention adds a node that retrieval systems cross-reference during answer generation.

An entity chain is the linked set of cross-domain references, structured data connections, earned media mentions, and knowledge graph entries that AI retrieval systems use to resolve, verify, and cite a brand. Each link in the chain is an independent verification point. When a retrieval-augmented generation (RAG) system encounters a query, it checks whether its knowledge graph can resolve the relevant entity with confidence by traversing these connections across multiple non-affiliated sources. Brands with complete entity chains across multiple domains get cited. Brands confined to a single domain are structurally invisible to AI retrieval systems regardless of content quality.

The degree to which a brand's identity is machine-readable — consistently resolved, attributed, and categorized by AI systems across the web. Layer 2 of the Machine Relations Stack.

Entity mass is the aggregate of explicit entity declarations, structured data, cross-domain references, and sameAs network density that AI retrieval systems evaluate when deciding which sources to cite. Unlike domain authority, which measures backlink equity, entity mass measures how clearly and consistently a brand's identity is declared across machine-readable surfaces. Google's Natural Language API scores entity salience on a 0 to 1 scale per document, but entity mass operates at the brand level: the sum of salience signals, structured data, and third-party entity references across every surface a retrieval system can access. Research across 500 B2B SaaS sites found a 0.71 correlation between structural entity factors and AI citation rates, compared to 0.18 for domain authority alone.

The distinction between three search optimization disciplines: SEO (Search Engine Optimization) targets ranked positions in traditional search results, AEO (Answer Engine Optimization) targets selection as the cited source in direct-answer features, and GEO (Generative Engine Optimization) targets citation inside AI-generated responses across ChatGPT, Perplexity, Google AI Mode, and other generative engines. The three disciplines share roughly 80% of their foundational work and diverge on the 20% that determines which surface cites you.

The cost structure, throughput dynamics, and resource allocation governing how trained AI models process queries at scale. A 1,000x cost collapse in three years made AI agents the default first reader of the internet, turning inference economics into the infrastructure foundation of Machine Relations strategy.

An AI system that decides which brands, products, and sources get recommended to users — the algorithmic successor to human gatekeepers like journalists and editors.

The category-defining discipline of earning AI engine citations and recommendations — the evolution from PR (convincing journalists) to MR (convincing machines like ChatGPT, Perplexity, and Gemini to cite and recommend your brand). Coined by Jaxon Parrott; AuthorityTech is the first agency to practice Machine Relations.

The process by which an AI engine moves from vague awareness of a brand to confident citation, the moment a machine transitions from recognizing a brand exists to recommending it in response to a specific user query.

Shorthand for Machine Relations — the evolution of public relations from convincing human journalists to convincing AI engines to cite, recommend, and surface your brand. Where PR 1.0 targeted media gatekeepers, PR 2.0 targets machine gatekeepers.

Public relations strategy designed to produce coverage that AI engines can retrieve, extract, and cite — prioritizing earned media authority, entity clarity, and machine-readable source architecture over raw press impressions.

Source credibility is the trust score AI engines assign to a content source when deciding what to cite in generated answers. Unlike traditional domain authority, which measures backlink equity, source credibility measures whether a source passes the content quality, author accountability, factual verification, and cross-domain corroboration checks that AI systems apply during retrieval and generation. A source can rank well in Google organic results and still fail source credibility evaluation because AI engines use a fundamentally different scoring model.

Topical authority is the degree to which search engines and AI retrieval systems recognize a website as the expert source on a specific subject. Unlike domain authority, which measures backlink strength across all topics, topical authority measures how deeply and consistently a site covers a defined subject area. Google's 2024 API leak confirmed it as a multi-signal system including siteFocusScore, siteRadius, and site-level topic embeddings. Research across 36 million AI Overviews found a 0.77 Spearman correlation between topic coverage breadth and AI citation likelihood, making topical authority the bridge between traditional search ranking and AI engine citation eligibility.

Earned media impact that occurs entirely within AI interfaces — brand awareness, trust, and demand generated when AI engines cite your brand in answers without the user ever clicking through to a source.

Any instance where an AI system names, recommends, or references a brand in a generated response — the foundational signal that determines whether a company exists in AI-mediated discovery.

A reference to a brand, product, or source within an AI-generated response — the fundamental unit of authority in machine-mediated discovery.

The references AI engines surface in generated answers, shaped by source trust, freshness, and cross-domain corroboration — and the emerging measurement discipline for tracking them.

The percentage of relevant AI queries that mention your brand in any context — the baseline visibility metric that establishes whether AI engines know you exist.

Website traffic sent directly from AI engines — ChatGPT, Perplexity, Claude, Gemini, and Google AI Overviews — when users click cited sources in generated answers. The fastest-growing acquisition channel in digital marketing, converting at 2-4x the rate of organic search.

The percentage of AI-generated responses in your category that mention or recommend your brand versus competitors — the AI-era metric for brand authority and market presence.

A brand's measurable presence across AI platforms — the composite metric that replaces impressions as the key Machine Relations performance indicator.

Unlinked text references to a brand name across the open web — news articles, publications, forums, YouTube, reviews, and directories — that AI retrieval systems consume as the primary signal for deciding which brands to cite.

The process of identifying where competitors are cited by AI engines and your brand is not — the AI-era equivalent of competitive keyword gap analysis.

Your brand's share of total category citations in AI-generated answers — the AI-era market share metric that reveals competitive position in machine-mediated discovery.

The rate at which AI citations for a brand increase over time — the compounding growth metric that separates sustained authority from one-time spikes.

The Machine Relations Index (MRI) is a public source-behavior dataset that tracks which root domains AI answer engines cite when responding to B2B buyer-intent questions. It classifies every observed source by deterministic source-role rules and reports source-segment citation rates and answer-engine breadth, publishing rates and rankings behind a public boundary that excludes query identifiers. The MRI was coined by Jaxon Parrott and is maintained as a public research standard at machinerelations.ai.

MRI Score is the Machine Relations Index metric for AI source authority. It reports how often AI answer engines cite each source domain as a source-segment citation rate, published only once a segment clears an evidence floor and graded into confidence tiers by how much evidence stands behind it.

The frequency at which AI engines actively recommend a brand when prompted with relevant purchase-intent queries — the highest-signal metric in AI-mediated discovery.

The degree to which a content section can be parsed, compressed, and reused by AI systems without losing its core meaning. The measurable property that determines whether retrieved content survives synthesis into AI-generated answers.

Content that delivers its core claim or answer in the opening sentences rather than after narrative buildup, maximizing the probability of AI extraction and citation.

A data point, quote, statistic, or named framework designed to be extracted and cited by AI engines in generated responses.

The practice of engineering content so AI systems can extract, verify, and cite specific claims, data points, and recommendations from it.

The practice of structuring a brand's digital identity so AI systems can resolve, verify, and cite it consistently across knowledge graphs and generated responses.

The process by which an AI system determines that multiple cross-web references — a LinkedIn company page, a news article mention, a schema.org markup block, a Wikidata entry — all point to the same real-world brand or person, and consolidates their authority signals into one entity rather than treating each as a separate, weaker source. Brands that resolve cleanly earn AI citations; brands that fail entity resolution get omitted or confused with competitors.

Structured data and digital presence markers that help AI engines identify, understand, and trust a brand as a distinct entity.

Content structured so AI engines can pull self-contained, verifiable claims from it — typically 40-60 word answer blocks with clear assertions and sourced statistics.

The set of technical, structural, and authority conditions a piece of content must meet before AI search engines will include it in their retrieval pool — the prerequisite layer that determines whether content can be cited at all.

An AI architecture that retrieves relevant documents from external sources before generating a response, grounding answers in real evidence rather than parametric memory alone.

The composite trust signal that determines whether an AI search engine cites a specific source when generating answers. Source authority is what separates a page that gets retrieved from a page that gets quoted.

The compounding advantage brands build when AI engines consistently cite them, where each citation reinforces the next and creates a self-widening competitive barrier.

The recency signal AI engines use to prioritize sources, where content updated within the last 90 days is cited at significantly higher rates and peak citation probability occurs within 7 days of publication.

AuthorityTech's five-layer strategic framework for AI visibility: Earned Authority, Entity Optimization, Citation Architecture, GEO/AEO execution, and Measurement.

The practice of optimizing for multiple AI platforms simultaneously, accounting for each engine's distinct citation preferences, retrieval methods, and authority signals.

A pay-per-placement public relations model that guarantees measurable media results, replacing the traditional retainer structure where brands pay monthly fees regardless of outcomes.

PR for AI search is the practice of earning third-party media coverage and expert citations that AI systems retrieve when generating answers — the earned authority layer of the Machine Relations stack that AuthorityTech operationalizes for B2B brands.

The cadence at which a brand publishes citation-grade content and earns media placements, directly correlated with the speed of AI visibility gains.

High-authority media outlets — Forbes, TechCrunch, WSJ, Reuters, The Guardian — that AI engines retrieve and cite at 3-5x the rate of mid-tier sources when constructing answers to buyer and category queries.

The emerging practice of AI agents autonomously evaluating, shortlisting, and selecting B2B vendors — Gartner predicts 90% of B2B purchasing will be intermediated by AI agents by 2028.

The measurable rate at which AI citation frequency drops when a brand stops producing fresh content — Perplexity citations erode within days; broader ChatGPT and Gemini visibility fades within weeks of publishing inactivity.

The narrowing period of opportunity to establish entity authority in AI engines before competitive citation patterns lock categories — closing 10x faster than prior platform shifts.

Buyer research activity occurring outside tracked channels — now predominantly in AI tools where 68% of B2B buyers begin vendor evaluation before any measurable touchpoint.

The winner-take-most dynamic in AI search where a small number of dominant brands capture a disproportionate share of AI citations — top 10 brands now capture 59.5% of category citations.

A firm specializing in generative engine optimization — structuring content, earning authority, and building entity signals so that AI search engines like ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude cite the client's brand in synthesized answers.

An AI-invented reference to a non-existent source, fabricated statistic, or misattributed claim — 31% of AI citations about B2B brands are hallucinated or materially inaccurate.

The structural split of the PR industry into traditional agencies in decline and Machine Relations-oriented agencies in growth — exemplified by Edelman's -8.1% revenue drop alongside Publicis's +4-5% gain.

A search where the user receives a complete answer directly in the results interface without clicking through to any website — now 69% of all searches, up from 56% pre-AI.

Sentiment Delta is the gap between how a brand intends AI systems to describe it and how those systems actually describe it in generated responses — a core Machine Relations measurement metric coined by Jaxon Parrott.

Generative Engine Optimization — the practice of structuring and distributing content so generative AI engines cite it in synthesized answers.

A Machine Relations metric, coined by Jaxon Parrott, that measures the percentage of AI-engine queries in which a brand is correctly identified, attributed, and represented as intended. Replaces share of voice as the primary AI-era brand measurement because volume of mentions means nothing if AI systems cannot confidently resolve which entity is being referenced.

The percentage of AI-generated answers that cite a brand, page, or source as a result of PR and earned media activity — the measurement that tells PR teams whether their placements are actually driving machine-mediated discovery.

The percentage of total AI engine citations your brand earns within a defined category or query set — the Machine Relations replacement for share of voice. Where traditional marketing measured brand presence across media impressions, share of citation measures brand presence across AI-generated answers in ChatGPT, Perplexity, Gemini, and similar engines.

The practice of structuring content to maximize visibility and citation in AI-generated answers across generative search engines like ChatGPT, Perplexity, Gemini, and Google AI Overviews — Layer 4 within the Machine Relations stack.

Google's AI-generated summaries that appear above traditional search results, synthesizing answers from multiple web sources with inline citations — the largest single surface where brands gain or lose visibility in AI-mediated discovery.