Glossary
The vocabulary of AI answers
The terms used across our docs, API reference and blog — the metrics are computable from the fields each monitor response returns.
- AI answer monitoring
- Repeatedly querying consumer AI products (ChatGPT, Perplexity, Gemini, Copilot, Grok) and Google's AI surfaces with the same prompts, then tracking how the answers change. The unit of measurement is the answer itself — text, sources and entities — not a ranking.
- Answer engine
- A product that returns a written answer to a question instead of a list of links: ChatGPT, Perplexity, Gemini, Copilot, Grok, Google's AI Mode and AI Overview. Also called generative engines.
- Answer rank
- The position of a brand or product in an AI answer's recommendation order — first named versus fourth. Measured from the order of entities[] in the response.
- AI Overview (AIO)
- Google's AI-generated answer block shown above organic results on some searches. Triggered per keyword and market; cited sources are the pages it read. See the AI Overview endpoint.
- AI Mode
- Google's standalone conversational search interface — follow-up questions, richer layouts, product and place blocks. A separate surface from AI Overview with different citation behavior.
- Citation
- A source an AI engine used to build its answer, shown as a numbered link. In the API, the sources[] array in result order. Being cited means your page was retrieved and used as evidence.
- Citation share
- Your domain's share of all citations across a prompt set: your citations divided by all citations returned. The AI analogue of SERP share of voice.
- Entities
- The brands, products and people an AI answer names, extracted to result.entities[]. The field used to compute mention rate and answer rank.
- Fan-out query
- A web search an AI engine runs behind the scenes before answering — usually several reworded or related versions of the user's prompt. Returned in searchQueries[]; the pages that rank for fan-out queries form the pool the answer cites from.
- GEO (generative engine optimization)
- The discipline of improving how often and how favorably AI answers mention and cite a brand. Measures mention rate and citation share rather than keyword position.
- Mention
- A brand or product named in an answer's text without necessarily being cited as a source. A mention without a citation means the model knows you through someone else's content.
- Mention rate
- The share of prompts in a monitored set whose answers name your brand. Tracked per engine and market; the top-line AI-visibility metric.
- Prompt set
- The fixed list of natural-language questions monitored on a schedule — the AI equivalent of a keyword list. Prompts are grouped by intent: category, comparison, problem, branded.
- Source position
- The order a cited page appears in sources[] — the number badge a user sees. Position 1 is the lead citation.
- Shopping card
- A product recommendation block inside an AI answer — name, price, rating, seller, link. Returned as shoppingCards[] (ChatGPT, Copilot) or products[] (AI Mode).
- Sync vs async
- A synchronous monitor call holds the request open and returns the answer in the response (base price plus a surcharge). An async task queues the work and delivers the result by webhook at base price — the right shape for volume.