GEO & AI Search Glossary

Generative Engine Optimization (GEO) is the practice of getting a brand cited, quoted or recommended inside AI-generated answers — in ChatGPT, Perplexity, Google AI Overviews, Gemini and Copilot — rather than ranking as a blue link. It's also called answer engine optimization (AEO) or LLM optimization (LLMO).

Related terms:AEO (Answer Engine Optimization)LLMO (Large Language Model Optimization)AI Share of VoiceAI Visibility Score

Answer Engine Optimization (AEO) describes the same goal as GEO — being the brand and source an AI answer engine cites — under a different name. In practice, GEO, AEO and LLMO are used interchangeably across the industry.

Related terms:GEO (Generative Engine Optimization)LLMO (Large Language Model Optimization)Answer Engine

Large Language Model Optimization (LLMO) is another name for the same discipline as GEO and AEO: optimizing content and brand presence so large language models like GPT, Claude or Gemini represent it accurately and recommend it.

Related terms:GEO (Generative Engine Optimization)AEO (Answer Engine Optimization)RAG (Retrieval-Augmented Generation)

An answer engine is a search interface that responds with a direct, synthesized answer instead of a ranked list of links — ChatGPT, Perplexity and Google's AI Overviews are answer engines. Winning visibility here means being the source the answer cites, not the top blue link.

Related terms:AEO (Answer Engine Optimization)AI OverviewAI ModeRAG (Retrieval-Augmented Generation)

Google AI Overview is an AI-generated summary Google shows above traditional search results for many queries, drawing on and citing a handful of web sources. Appearing as one of those cited sources is a core GEO objective.

Related terms:AI ModeAnswer EngineSchema Markup

Google AI Mode is Google's fully conversational search experience — a chat-style interface for multi-turn queries, distinct from the AI Overview panel shown above classic results. Aurora tracks visibility in both separately, since they surface different sources.

Related terms:AI OverviewAnswer EngineGEO (Generative Engine Optimization)

AI share of voice measures how often AI assistants recommend a brand for the prompts that matter to it, relative to its competitors. Because a single AI answer only names a handful of brands, share of voice is a sharper signal than a ranking position: either the brand is in the answer, or it isn't.

Related terms:AI Visibility ScoreCitation RateGEO (Generative Engine Optimization)

Citation rate is the share of tracked AI responses that cite a brand's own domain as a source, as opposed to only mentioning the brand's name in passing text. A high mention count with a low citation rate usually means the AI recognizes the brand but isn't treating its own content as authoritative for the topic.

Related terms:AI Share of VoiceRAG (Retrieval-Augmented Generation)E-E-A-T

An AI visibility score condenses how often, how prominently and how favorably a brand appears across tracked AI engines into a single trackable number, so changes in visibility over time are easy to see at a glance.

Related terms:AI Share of VoiceCitation RatePrompt Intelligence

Prompt intelligence is the analysis of the real questions people actually ask AI assistants about a category or brand — rather than guessing at keywords the way traditional SEO does. It shows which prompts a brand wins, loses, or never appears for.

Related terms:AI Visibility ScoreKeywordGEO (Generative Engine Optimization)

Retrieval-Augmented Generation (RAG) is the technique behind most modern AI search: the system first retrieves relevant documents or web pages, then generates its answer grounded in that material — which is why the sources an AI cites are traceable to real, current content rather than only its training data.

Related terms:Answer EngineCitation RateCrawling

Schema markup is structured data, usually in JSON-LD format, embedded in a page's code to describe its content in a machine-readable way — telling search engines and AI crawlers exactly what an organization, product or article is, rather than leaving them to infer it from prose.

Related terms:Technical SEOE-E-A-TAI Overview

Core Web Vitals are Google's user-experience metrics — loading performance, interactivity and visual stability — that factor into classic search ranking and, indirectly, into how reliably AI crawlers can fetch and parse a page.

Related terms:PageSpeedTechnical SEO

Crawling is the process by which a search engine or AI system's bots systematically visit and read web pages. A page that can't be crawled can't be cited — which is why website optimization tools include crawler-accessibility checks.

Related terms:IndexingRobots.txtSitemap

Indexing is what happens after crawling: a search engine stores and organizes a page's content so it can be retrieved later for relevant queries. Being indexed is a prerequisite for being cited by an AI answer engine — no engine can quote a page it has never indexed.

Related terms:CrawlingCanonical URLMobile-First Indexing

PageSpeed measures how quickly a page loads and becomes usable. Slow pages hurt both classic rankings and the reliability of AI crawlers trying to fetch and parse content within their own time budgets.

Related terms:Core Web Vitals

Mobile-first indexing means Google primarily uses a site's mobile version — not its desktop version — to index and rank content. A page that renders poorly on mobile is effectively the version most crawlers, human and AI, actually see.

Related terms:Indexing

The canonical URL tells search engines which version of a page is the ‘real’ one when duplicate or near-duplicate content exists at multiple URLs, consolidating ranking (and citation) signals onto a single address instead of splitting them.

Related terms:IndexingKeyword Cannibalization

HTTPS is the encrypted version of the HTTP protocol. It has been a baseline ranking signal for years, and browsers now flag non-HTTPS sites as insecure — a page an AI crawler can't fetch securely is a page it's less likely to trust as a citation.

Related terms:Technical SEO

Robots.txt is a plain-text file at a site's root that tells crawlers — search engines and AI bots alike — which parts of the site they may or may not access. Blocking an AI crawler here means that content can never be cited, however good it is.

Related terms:CrawlingSitemapAgent-Readiness (AAO)

An XML sitemap lists a website's important URLs, helping search engines and AI crawlers discover and prioritize what to crawl — especially useful for large sites or pages with few internal links pointing to them.

Related terms:CrawlingIndexingRobots.txt

Technical SEO is the discipline of making a site crawlable, indexable and fast — the infrastructure layer underneath content quality. Every AI crawler capability Aurora checks (crawlability, structured data, page speed) falls under technical SEO.

Related terms:CrawlingIndexingCore Web VitalsSchema Markup

E-E-A-T — Experience, Expertise, Authoritativeness, Trustworthiness — is Google's framework for judging content quality. AI answer engines built on similar retrieval and ranking signals tend to favor the same qualities when choosing which sources to cite.

Related terms:Citation RateDomain AuthorityBacklink

A SERP (Search Engine Results Page) is the page of results a search engine returns for a query. AI Overviews and AI Mode panels increasingly appear directly on, or instead of, the classic link-based SERP.

A backlink is a link from another website pointing to yours. It remains one of the strongest traditional ranking signals, and the same authoritative pages that earn backlinks are often the ones AI answer engines choose to cite.

Link building is the deliberate practice of earning backlinks from other websites, historically one of the most resource-intensive SEO disciplines. The same authoritative links that link building targets are often the ones AI answer engines treat as trust signals.

A keyword is the search term or phrase content is optimized to match. GEO shifts the emphasis from matching keywords to matching the actual questions people ask AI assistants — prompt intelligence, not keyword research.

Keyword cannibalization happens when multiple pages on the same site compete for the same search term, splitting ranking signals between them instead of consolidating authority on one page.

The meta description is the short summary shown under a page's title in classic search results. It doesn't directly influence rankings, but a clear, accurate one improves click-through — and gives an AI crawler a concise, human-written summary of what the page is about.

Alt text is the descriptive text attribute on an image, read by screen readers and search engines that can't ‘see’ the image itself. AI crawlers rely on the same text when an answer engine needs to understand or cite an image-heavy page.

Internal linking connects pages within the same website, helping both users and crawlers understand site structure and find related content. Well-linked pages are easier for AI crawlers to discover and contextualize.

Domain authority is a third-party score (popularized by Moz) estimating how likely a whole domain is to rank, based on its backlink profile. It's not a Google metric, but it correlates with the kind of established, trusted domains AI answer engines also tend to cite.

Bounce rate is the share of visitors who leave a page after viewing only that one page, with no further interaction. It's a classic engagement signal — a high bounce rate on a page an AI engine sends traffic to still hurts conversion, even if the citation itself succeeded.

Local SEO optimizes a business's visibility for geographically specific searches — ‘near me’ queries and map results. AI assistants answering local questions draw on the same local signals: Google Business Profile, reviews, and citations.

A Google Business Profile is a free, verified business listing that powers a company's presence in Google Maps and local search results — foundational for local SEO, and increasingly a source AI assistants draw on for local business questions.

On-page SEO covers everything controlled directly on a page: content, headings, meta tags, internal links and structure. It's the foundation GEO builds on — you can't be cited accurately if the underlying content isn't clear in the first place.

Off-page SEO covers signals earned outside your own website — backlinks, mentions, reviews, social signals and overall reputation. AI answer engines weigh many of the same off-page trust signals when deciding which sources to cite.

The agentic web describes the shift from a web browsed by humans to one increasingly navigated by autonomous AI agents — researching, comparing and even transacting on a user's behalf. It raises a new visibility question: is a brand findable and correctly represented not just to people, but to agents acting for them?

Related terms:MCP (Model Context Protocol)A2A (Agent2Agent)Agent-Readiness (AAO)

The Model Context Protocol (MCP) is an open standard, introduced by Anthropic, that lets AI agents connect to external tools and data sources in a consistent way. It has become one of the most widely adopted standards for wiring an agent up to real applications.

Related terms:Agentic WebA2A (Agent2Agent)

Agent2Agent (A2A) is an open protocol for communication between AI agents, originally developed by Google and now maintained by the Linux Foundation, built on JSON-RPC.

Related terms:Agentic WebMCP (Model Context Protocol)

Agent-readiness is the set of measures that let autonomous AI agents find, understand and act on behalf of a business through open standards like A2A and MCP — the agentic-web counterpart to making a site readable by AI answer engines.

Related terms:Agentic WebRobots.txtSchema Markup