The complete glossary of AI search, GEO, AEO and SEO in 2026
50+ terms defined for the new era of generative search. Updated May 2026 to reflect ChatGPT-5 with browse, Google AI Overviews global rollout, Perplexity’s Comet browser launch, and the consolidation of GEO/AEO as standalone disciplines distinct from classical SEO.
Run a free AI visibility audit → Pricing
Core concepts
SEO (Search Engine Optimization)
The discipline of optimizing web pages to rank highly in classical search engine results pages (SERPs). Born in the late 1990s. Primary engine: Google. Primary signals: backlinks, on-page content, technical performance, intent match. Still relevant in 2026 but no longer sufficient.
GEO (Generative Engine Optimization)
The discipline of optimizing content to be cited and recommended inside AI-generated answers. Engines: ChatGPT, Perplexity, Google Gemini, Anthropic Claude, Microsoft Copilot. Primary signals: structured data, entity clarity, factual density, third-party authority, recency.
AEO (Answer Engine Optimization)
Often used interchangeably with GEO. Some practitioners distinguish: AEO = optimizing for direct answer features (Featured Snippets, AI Overviews); GEO = broader optimization for any generative output. We treat them as overlapping with shared methods.
LLM (Large Language Model)
Neural networks trained on massive text corpora that generate human-like text. Examples: GPT-5 (OpenAI), Claude 3.5 (Anthropic), Gemini 2 (Google), Llama 3 (Meta). LLMs power all major AI search engines in 2026.
RAG (Retrieval-Augmented Generation)
The technique most modern AI engines use to ground responses: instead of relying purely on training data, the model first retrieves relevant documents from the live web (or a knowledge base), then synthesizes an answer citing those sources. Why your fresh, well-structured pages can be cited.
AI Overview (AIO)
Google’s AI-generated answer that appears at the top of search results pages for many queries since 2024. Synthesizes information from multiple web sources. Cites typically 3-4 URLs. Replaces the need to click through to traditional results for ~30% of informational queries.
Featured Snippet
A direct extract from a single web page that Google displays at position zero of the SERP. Predates AI Overviews. Different mechanism: extracts an existing passage rather than synthesizing new text. Still active in 2026 alongside AI Overviews.
AI engines and platforms
ChatGPT
OpenAI’s conversational AI, 800M weekly active users in March 2026. Most-cited generative engine for brand discovery in 2026. With browse mode (default since GPT-5), actively cites web sources.
Perplexity
AI search engine specialized in citing sources. Highest source density per response (mean 6.2 sources). 56M weekly users. Strong recency bias (favors content updated in last 90 days).
Google Gemini
Google’s LLM, integrated across Search (AI Overviews), Workspace, Android. Mixes training data with grounded sources. Conservative citer compared to Perplexity.
Claude
Anthropic’s LLM, used both as consumer chatbot and via API by businesses. Most conservative citer (mean 2.1 sources). When Claude cites you, it considers you a primary source.
Microsoft Copilot
Microsoft’s AI assistant, powered by GPT-4o + Bing search. Integrated in Windows, Office, Edge. Primarily B2B/productivity audience.
Apple Intelligence
Apple’s on-device AI layer (iOS 18+, macOS 15+). Mostly local processing, with optional ChatGPT and Gemini fallback. Expected to grow as a discovery channel through 2027.
Optimization signals
Schema.org
The vocabulary used to add structured data to web pages, helping crawlers understand content. Critical for GEO. Top-priority schemas in 2026: Organization, LocalBusiness, Hotel, Restaurant, Product, FAQPage, HowTo, Article, BreadcrumbList.
FAQ Schema
Specific schema (FAQPage) marking up question-answer pairs. Adds +90% citation lift on average across CapstonAI customer pages. The cheapest, fastest schema win in 2026.
JSON-LD
The recommended format for embedding Schema.org data in HTML pages. Sits in a `<script type=”application/ld+json”>` tag. Easier to maintain than microdata or RDFa.
Wikidata
Wikipedia’s structured-data sister project. The single most-cited entity database by all four major LLMs. Brands with verified Wikidata entries are cited 3.1× more (CapstonAI Q1 2026 benchmark).
Knowledge Graph
Google’s database of entities and their relationships. Powers the right-side info box on search results. Major source signal for AI Overviews. Tied to Wikidata + Wikipedia for most entities.
Entity
A unique, identifiable thing (brand, person, place, product, concept). LLMs reason about entities, not just keywords. “Entity SEO” or “entity-based optimization” focuses on making your brand legible as a coherent entity to AI engines.
NAP
Name, Address, Phone. Consistency across all platforms (Google Business Profile, Booking, TripAdvisor, Yelp, etc.) is a foundational signal for local SEO and GEO.
Citation (in AI search context)
When an AI engine references your brand, content, or URL inside its generated response. Different from a backlink: a citation can exist without any clickable link. Measured via prompt panel testing (see AI citation tracking).
Share of Voice (AI)
Your citation rate vs competitors across a defined prompt panel. Calculated per engine, per prompt category, per market. The core KPI of GEO programs.
Performance and crawlability
Core Web Vitals (CWV)
Google’s set of user experience metrics: LCP, INP, CLS. Confirmed ranking signals across Google Search and AI Overviews since 2021. Slow pages get partial parsing or deprioritization by AI crawlers.
LCP (Largest Contentful Paint)
Time to render the largest visible element on a page. Image-driven on 71% of pages. Should be under 2.5 seconds. WebP image conversion is the cheapest, fastest LCP win.
WebP
Modern image format from Google. Average 26% smaller than JPEG/PNG at equal quality. 97.4% browser support in May 2026. See WebP optimization.
AVIF
Newer image format with even better compression than WebP. 94.6% browser support in May 2026 (including Safari 16+). Recommended as a complement to WebP, not a replacement.
Crawler / AI Bot
Automated agents that fetch web pages on behalf of search and AI engines. Major AI crawlers in 2026: GPTBot, ChatGPT-User, ClaudeBot, PerplexityBot, Google-Extended, Common Crawl. Most respect robots.txt and have per-page time/bandwidth budgets.
robots.txt
The standard file telling crawlers which parts of your site they may or may not access. Critical decision in 2026: do you allow GPTBot, Google-Extended, ClaudeBot? Allowing increases AI visibility but may reduce control over how your content is reused.
Content patterns
Pillar page
A long-form (1 500-3 000+ words) authoritative page covering a broad topic, linking out to focused “child” pages. Hub-and-spoke architecture. LLMs love pillar pages because they’re entity-clear and structurally rich.
Silo
Architectural pattern: one parent topic page linked tightly to a set of related child pages, with controlled cross-linking. Reinforces topical authority. Recommended for multi-location, multi-product, or multi-vertical brands.
Comparison content
“X vs Y”, “alternatives to Z”, “best [category] for [buyer]” formats. Cited 2.8× more than generic feature pages by AI engines. The single highest-leverage content type for GEO in 2026.
Listicle
Top 10, Top 25, Top 100 format pages. Cited 1.7× more than narrative pages. LLMs extract list items as discrete answer candidates.
Authority signals
Backlink
An external link from another site to yours. Classical SEO signal. Still relevant in 2026 but less central for GEO than schema, entity clarity, and Wikipedia/Wikidata presence.
Domain Authority (DA) / Domain Rating (DR)
Third-party metrics (Moz, Ahrefs) approximating overall domain strength based on backlink profile. Useful directional indicators, not Google or AI engine signals directly.
E-E-A-T
Google’s Experience, Expertise, Authoritativeness, Trustworthiness framework. Used by quality raters. Strong correlation with AI Overview eligibility. Demonstrated through About pages, author bios, sources cited.
Wikipedia
The #1 source cited by all four major LLMs. If your brand is notable enough to have a Wikipedia article, having one is the single highest-leverage authority move.
Measurement
AI citation tracking
The discipline of measuring how often, where, and how favorably AI engines cite your brand. Different from brand monitoring (open web) or SERP tracking (Google blue links). See AI citation tracking.
Prompt panel
The defined set of prompts used to systematically query AI engines on behalf of a tracked brand. Typical sizes: 20 prompts (small B2B) to 100+ (multi-market enterprise). Updated quarterly to reflect evolving search behaviors.
Sentiment (in AI citations)
The tone of mentions in AI responses: positive, neutral, or negative. Tracked alongside citation rate. Negative sentiment in citations is rare but high-impact when it occurs.
Traffic and attribution
AI traffic / Generative traffic
Sessions originating from AI engines, identified via referrer (chat.openai.com, perplexity.ai, gemini.google.com) or via dedicated tracking parameters. Estimated at 22% of branded discovery traffic for B2B/ecommerce in 2026.
Zero-click search
Searches where the user gets the answer directly in the SERP (Featured Snippet, AI Overview, Knowledge Panel) without clicking any result. Growing share of search behavior since 2020, accelerated by AI Overviews.
Brand discovery traffic
Visits originating from queries that include your brand name or are clearly aimed at finding/evaluating you. AI engines now drive 22% de cela pour mid-market brands (Gartner, March 2026).
Bots, blocking, and content protection
GPTBot
OpenAI’s crawler used to gather content for ChatGPT training and live retrieval. Can be allowed/blocked via robots.txt. Allowing it grants OpenAI permission to use your content for training.
ChatGPT-User
OpenAI’s user-driven crawler — fetches a page when a ChatGPT user asks about it. Less aggressive than GPTBot but should be allowed for live citation eligibility.
Google-Extended
Google’s controllable crawler for Bard/Gemini training. Can be blocked separately from regular Googlebot, allowing you to keep search ranking while opting out of AI training.
Adjacent disciplines
Semantic SEO
Optimization based on topic and entity relationships rather than just keywords. Foundational layer for GEO. Tools: TopicMap, Surfer SEO, Frase, Clearscope.
International SEO / GEO
Optimizing for multiple countries and languages. Hreflang, market-specific schemas, local entity presence. AI engines have notably different citation behavior across languages — measured optimization per market is mandatory.
Local SEO
Optimization for “near me” and place-based queries. Google Business Profile is the primary asset. AI engines now factor in GBP signals heavily for hospitality, retail, and services.
Related reading
- AI citation tracking
- AEO Insights
- What is GEO
- AEO vs SEO
- Best AI SEO tools 2026
- WordPress AI SEO plugin
- Shopify AI SEO app
Need to apply this glossary to your brand?
Run a free AI visibility audit →
Last updated: May 2026. Compiled by CapstonAI from internal benchmarks (24 800 LLM responses Q1 2026), public engine documentation, Schema.org spec, Web.dev Core Web Vitals docs, caniuse browser support data.