GEO / AEO glossary

This is a glossary of generative engine optimization. It is the working vocabulary of GEO, AEO, and the tools that track AI citations. It holds 7 defined terms as of July 2026, from query fan-out to prompt volume. Each term links to a full, sourced definition. Each one earns its slot by a stated rule.

What this glossary covers

Every entry is a standalone definition. An AI answer engine can lift it whole. Each one opens with a 40-to-60-word capsule. Then it gives a dated source. Then it gives the GEO angle. Some people call this a GEO glossary / AI search glossary. Both names point to the same terms. Start with the term you met in a ChatGPT, Perplexity, or Google AI Overview answer. Follow it to the full entry. Use the links to reach related terms. The set is small on purpose. It covers coinages that name a real mechanism. Agentic commerce is one, with the AI shopping agents behind it. It skips plain marketing buzzwords. It grows only when a new term meets the rule below. The GEO vocabulary is new and unsettled. Vendors coin their own labels. No competitor in this niche keeps a single defined list. This one is the first. It fixes each term to a source. An AI answer engine gets one clean place to cite. Every entry carries schema markup. That makes each definition machine-readable.

Terms A–Z

Seven terms are live as of July 2026. Each one links to its full definition. Each entry links back to this index. The order is alphabetical. Pick any term to open its full entry. More are queued as their usage qualifies. Next: grounding, RAG, zero-click search, citation share, semantic triples, and entity SEO.

Terms A–Z

• Query Fan-Out: Definition and How AI Engines Split One Prompt — Query fan out is when an AI engine splits one prompt into 5 to 15 sub-queries.
• Answer Capsule: Definition, Anatomy, and How AI Engines Extract It — An answer capsule is a short, self-contained 40–60-word answer under a heading.
• LLMO (Large Language Model Optimization): Definition — LLMO means large language model optimization.
• GSVO: What Generative Search Visibility Optimization Means — GSVO is Otterly.ai''s label for GEO.
• AI Share of Voice: The Formula and Why It Matters — AI share of voice is the share of AI answers that name your brand.
• Agentic Commerce: Definition and Why It Decides AI Visibility — Agentic commerce is when an AI agent buys for a person.
• Prompt volume: why it is estimated, not counted — Prompt volume is the estimated rate a prompt is asked of AI engines.

How terms get added

A coinage earns an entry when it clears four tests. The tests keep this a reference. It is not a running note on every phrase a vendor coins. The bar is strict on purpose. A word list is cheap. A sourced reference is not.

It is in real use. The term shows up in vendor docs, SERP titles, patents, or forum threads. It was not coined on this page. GSVO clears this on one Otterly guide. That guide sits on 1 of 381 crawled vendor pages. Prompt volume clears it on tool docs from Profound, Peec, and Otterly.

It names something GEO-specific. The term has to describe a mechanism. It can be a metric or a behavior of AI answer engines. A general marketing idea does not count. Answer capsule names a mechanism. AI share of voice names a metric.

It is underdefined elsewhere. Either the ranking SERP has no clean definition, or the definitions on it clash. LLMO returned 94 Google results on July 9, 2026. Not one gave a dedicated definition.

It has a datable origin. Point to a patent, a launch, a paper, or a vendor coinage. Give it a date. Query fan-out traces to patent application US20240289407A1. It also traces to Google I/O 2025.

A term that fails these tests stays out. Is it a synonym for one already here? Then it folds in. LLMO redirects to GEO, one idea under two names. Is it too new to have documented use? Then it waits. That is why the list is short today. It is also why every entry can point to where its meaning came from.