The AI Search Glossary

Canonical definitions for every term used across Astiva's product, methodology, and reports. 30 terms with fact-first definitions and related-term cross-links.

Core Category Terms

AISO (AI Search Optimization) — Monitoring, analyzing, and improving how a brand appears in AI-generated answers across ChatGPT, Claude, Gemini, Perplexity, and other AI platforms.

AEO (Answer Engine Optimization) — Optimizing content so answer-oriented surfaces (Google AI Overviews, Bing Copilot) extract it as the direct answer.

GEO (Generative Engine Optimization) — Optimizing content so generative AI engines cite, recommend, or mention a brand inside responses. Overlaps with AISO.

The 7 AISO Metrics

Visibility Percentage — share of tracked prompts mentioning the brand.

Share of Voice — brand mentions / (brand + competitor mentions).

Average Position — where the brand appears inside an AI response (first, middle, later).

Brand Sentiment — tone of each mention, aggregated.

First Mention Rate — share of responses naming the brand before any competitor.

Mention Frequency — total brand mentions per time window.

Sentiment Volatility — week-over-week variance in sentiment; flags emerging issues.

Technical Terms

Brand Normalization — resolving brand mentions across casing, spacing, hyphenation, and misspellings.

Citation — AI platform reference to a specific source in a generated response.

Citation Gap — prompt where a competitor is cited and the brand is not.

Citation Velocity — rate at which a brand earns new AI citations over time.

Knowledge Cutoff — date through which an AI model was trained.

RAG (Retrieval-Augmented Generation) — AI architecture that fetches external info at query time.

Hallucination — confidently expressed AI statement that is factually incorrect.

Query Sampling — methodology step that builds a platform-specific query set across intent buckets.

Prompt Enrichment — expanding a base prompt with context before sending to an AI model.

Structured Data & Discovery

llms.txt — emerging convention for a root-level text file describing a site for AI crawlers.

llms-full.txt — longer companion to llms.txt with full product, pricing, methodology, and FAQ payload.

JSON-LD — format for embedding Schema.org structured data in HTML.

Schema.org — shared vocabulary of structured-data types (Organization, FAQPage, Product, etc.).

FAQPage Schema — Schema.org type wrapping Q&A pairs; lifts featured-snippet and AI Overview extraction.

Platforms & Surfaces

AI Overview — Google-generated AI summary at the top of some search results.

AI Mode — Google's conversational search experience.

Zero-Click Search — search where the user gets the answer on the results surface without clicking through.

Featured Snippet — highlighted extract at the top of Google results that directly answers a query.

E-E-A-T — Experience, Expertise, Authoritativeness, Trustworthiness; Google's content-quality framework, increasingly weighted by AI search.

Share of Model — across AI models a buyer might use, how many name this brand in the decision moment.

Mention Extraction — pipeline step that parses an AI response for mentions, position, sentiment, and competitors.

Full glossary | See the methodology

← Back to Astiva AI