Glossary
Glossary
Learn the meaning of key AI search and marketing terms in Gravton’s easy-to-navigate glossary, designed to help you understand AI visibility, AEO, and GEO faster.
Haritha Kadapa
Glossary
The Rise of AI Search
AI Overviews: AI-generated summaries shown in Google search results that synthesize information from multiple sources to directly answer a user’s query.
AI Prompts: The questions or instructions users submit to AI models to generate responses. AI prompts are typically longer, conversational, and more context-rich than traditional search queries.
AI Recommendations: A type of AI-generated output in which a platform names specific brands in response to a user seeking guidance.
AI Search: A form of search where artificial intelligence systems generate direct answers to user questions by synthesizing information from multiple sources rather than returning a list of links.
Large Language Models (LLMs): AI models trained on large datasets to generate human-like text and responses. Examples include models used by platforms such as ChatGPT, Gemini, and Claude.
Retrieval Phase: The stage of LLM response generation in which the model actively retrieves sources that appear credible, relevant, and consistent with the query, effectively conducting a real-time signal assessment. The retrieval phase complements the training phase and directly influences which brands and descriptions appear in a given AI response.
Training Phase: The stage of LLM development in which models ingest large text corpora and learn statistical associations between brand names, descriptors, and contexts. Patterns established during training form the baseline of how a model characterises a brand before any retrieval occurs.
Zero-Click Search: A search interaction in which the user's query is fully resolved by an AI-generated summary or featured snippet, resulting in no click to an external website.
Generative Engine Optimization (GEO) & Answer Engine Optimization (AEO)
Answer Engine Optimisation (AEO): The practice of structuring and formatting content so that it appears in direct answer boxes and Google AI Overviews.
Content Gap Analysis: The process of identifying queries where a competitor is cited in AI responses and a brand is not, then determining what content type is responsible for the gap, such as structured definitions, FAQ pages, comparison tables, or original research.
Generative Engine Optimization (GEO): The practice of optimizing digital content so AI platforms such as ChatGPT, Perplexity, and Google AI Overviews can discover, understand, and cite it in their generated responses.
Intent Intelligence: The practice of understanding the meaning, motivation, and goal behind user prompts rather than just the keywords used. It helps align content with what users are actually trying to accomplish in AI search.
Opportunity Gap: A prompt or query cluster where user intent and search demand are strong but a brand's content coverage is weak, unstructured, or absent entirely. Opportunity gaps represent the highest-impact areas for content investment in an AI visibility strategy.
Prompt Market Analysis: The practice of collecting and studying large volumes of user prompts across AI platforms to identify patterns, emerging demand, and competitive visibility. Where intent intelligence explains the why behind an individual prompt, prompt market analysis answers the what at scale, revealing which questions buyers are asking, how they phrase them, and which brands are being cited in responses.
Prompt Taxonomy: The practice of classifying prompts by intent type and buying journey stage to prioritise which queries matter most for content strategy.
Topical Authority: The depth and breadth of a website's content coverage within a specific subject area, used by AI models to identify reliable domain experts and prioritize sources during retrieval.
AI Visibility & Brand Presence
AI Mention Rate: The percentage of relevant queries in which a brand appears within AI responses, used as a primary metric for measuring AI narrative performance.
AI Search Visibility: A metric that measures how frequently a brand, product, or website appears in AI-generated answers across platforms like ChatGPT, Perplexity, and Google AI Overviews.
AI Share of Voice: The percentage of relevant AI responses in which a brand appears, measured relative to competitors across a defined set of target queries.
AI Visibility: The degree to which a brand appears, is accurately described, and is positively represented in AI responses.
Citation Share: The percentage of relevant prompts in which a brand is named, or its domain is cited within an AI response.
Positioning Rate: A metric that measures how often a brand is named as a primary recommendation versus merely mentioned as a secondary or background option in AI responses.
Prompt Coverage Rate: The proportion of a defined prompt framework in which a brand has any presence at all across AI platforms. It measures the breadth of visibility rather than frequency, showing how many distinct queries a brand is answering versus those it is missing.
Share of AI Voice: A metric measuring a brand’s presence in AI responses relative to its competitors across relevant queries.
AI Brand Monitoring & Narrative Analysis
AI Brand Monitoring: The process of tracking how AI platforms mention, describe, and evaluate a brand. It goes beyond traditional monitoring by analyzing AI-generated summaries, recommendations, and comparisons.
AI Narrative: The synthesized description that AI platforms generate about a brand, combining signals like credibility, citation frequency, and topical authority to form a summarized “story” about a company.
AI Narrative Audit: A periodic assessment that captures how AI models currently describe a brand, typically conducted by testing queries across multiple AI platforms and analyzing sentiment and positioning.
AI Sentiment Analysis: The process of evaluating how AI models describe a brand by classifying the tone of generated responses as positive, negative, neutral, or mixed. It focuses on the tone and positioning within AI summaries and recommendations.
Narrative Drift: The measurable gap between a brand’s intended positioning and how AI platforms actually describe it. For example, a brand that claims to be an “industry leader” but is described by AI as a “niche provider.”
Sentiment Distribution: The breakdown of AI-generated brand descriptions across positive, neutral, negative, and mixed categories, expressed as a percentage across a defined query set.
Sentiment Gap: The measurable difference between how a brand is described in AI responses compared to its competitors within the same responses.
Buyer Intent & the AI Funnel
Bottom of Funnel (BOFU): The decision stage where users are ready to take action, such as purchasing a product, signing up for a service, or requesting a demo.
Commercial Investigation Intent: A buyer journey stage in which the user is actively comparing options and evaluating which solution best fits their needs. Prompts at this stage, such as "Brand Z vs Brand Y," carry significant AI influence because models frequently generate comparison responses here.
Informational Intent: A buyer journey stage in which the user wants to learn something about a category or topic.
Intent Category Coverage: The distribution of a brand's AI presence across prompt types: informational, navigational, commercial investigation, and transactional, is used to identify which intent stages a brand is visible in and which represent gaps.
Invisible Funnel: The portion of the buyer journey that occurs within AI platforms, including research, comparisons, and brand evaluation.
Middle of Funnel (MOFU): The evaluation stage where users compare products, services, or solutions to determine which best meets their needs.
Top of Funnel (TOFU): The awareness stage of the marketing funnel where users seek general information to understand a problem or topic.
Transactional Intent: A buyer journey stage in which the user is ready to take a specific action, such as signing up, booking a demo, or downloading a resource.
Citations, Authority & Trust Signals
Citation Frequency: The number of times a brand is mentioned or cited across a defined set of AI platform queries over a given period.
Citation Graph: A network of nodes (content sources) and edges (citations or links) that maps how web content references other content. AI models use it to determine which sources are credible, well-connected, and worth citing.
Citation Quality: A measure of the authority of the sources AI platforms draws upon when referencing a brand. High citation quality means AI systems are pulling from credible, high-authority sources; low citation quality means the brand is cited primarily from low-authority or outdated content.
Core Nodes: High-authority sources within a citation graph, such as peer-reviewed publications, major industry outlets, and government data, that are frequently referenced across multiple domains.
Domain Authority: A measure of how consistently authoritative sources reference a given domain, used by AI models as a key ranking signal when selecting sources to cite.
Experience-Expertise-Authoritativeness-Trustworthiness (E-E-A-T): The criteria AI models and search engines use to evaluate whether a source is credible enough to cite or rank.
Mid-Tier Nodes: Content sources cited by core nodes, such as industry publications or expert-run blogs, that carry significant weight when embedded in the right citation clusters.
Peripheral Nodes: Isolated pages with few or no inbound references from authoritative sources, which AI models assign minimal influence regardless of content quality.
Semantic Consistency: The degree to which a source is reliably associated with a given topic across multiple documents, used by AI models to identify domain experts during retrieval.
Source Influence: The degree to which a content source shapes AI responses, determined by citation frequency, domain credibility, and semantic consistency across sources.
Technical GEO Foundations
Canonicalization: The practice of specifying the preferred URL version of a page using a canonical tag, ensuring that AI models and search engines reference the authoritative source when multiple URLs contain similar content.
FAQPage Schema: A structured data markup type that formats content as explicit question-and-answer pairs in machine-readable JSON-LD.
JSON-LD: JSON-LD (JavaScript Object Notation for Linked Data) is the recommended format for implementing schema markup on web pages. JSON-LD is placed in the <head> section of a page and allows AI crawlers to access structured metadata without relying on visible page content or JavaScript rendering.
llms.txt: A plain-text file placed at the root of a domain that guides AI models on which pages to prioritise when crawling and indexing content. It is an emerging convention in technical GEO used alongside robots.txt to improve AI crawlability and direct AI systems toward high-value pages.
Schema Markup: Structured data code added to a webpage that helps search engines and AI models accurately identify and extract content such as FAQs, instructions, and organization details.
Server-Side Rendering (SSR): A web architecture pattern in which page content is generated on the server and delivered as complete HTML before reaching the browser. SSR is critical for AI crawlability because most AI crawlers do not reliably execute client-side JavaScript, meaning content loaded dynamically may be invisible to them.
User Agent: A text string sent by a browser, application, or bot to a web server that identifies the software making the request. It also provides information about the device, operating system, and browser being used.
Attribution, Analytics & ROI
AI Attribution: The process of tracking how AI responses influence user behavior and contribute to awareness, consideration, and purchasing decisions.
AI-Influenced Pipeline: The set of sales opportunities that interacted with an AI platform at some point in the buyer journey, even if the last recorded touchpoint was branded search or direct traffic. It is a key input in the GEO ROI formula and is calculated either through self-reported source fields in a CRM or by modelling branded search lift.
Assisted Conversion: A deal or signup in which an LLM interaction occurred somewhere in the buyer journey, even if the last recorded click was branded search or direct traffic. Assisted conversions are tracked using CRM self-reported attribution fields and multi-touch attribution models.
Customer Data Platform (CDP): A system that unifies customer data from multiple sources, behavioural, transactional, and identity into a single profile. In the context of AI search, a CDP connects anonymous AI-referred website visits to known customer profiles, enabling teams to understand how AI discovery contributes to pipeline and revenue.
Data Pipeline: The automated process that moves, transforms, and loads data from source systems into a centralised repository for analysis. An AI-ready data pipeline must go beyond click-based signals to ingest non-click inputs such as citation frequency, share of AI voice, and branded search lift.
Direct LLM Traffic: Website sessions that originate from a user clicking a link directly within a ChatGPT, Perplexity, or similar AI platform response.
GEO ROI: The return on investment from a Generative Engine Optimisation programme, calculated as: (influenced pipeline × close rate × average deal size − programme cost) ÷ programme cost. GEO influence often precedes trackable interactions, so the influenced pipeline figure is derived from a blend of directly reported and modelled attribution.
UTM Parameters: Tracking codes added to the end of a URL that enable analytics tools to identify the source, medium, and campaign of a website visit.
VISIBILITY & CONTENT STRATEGY
