How we define terms
Every definition in this glossary is written by the Quratic editorial team and, where we cite a figure, backed by data from our own measurement network — not paraphrased from other glossaries.
How definitions are written
Each entry leads with a self-contained 40–60 word definition, followed by how the term works, how it differs from adjacent terms, a concrete example, and — where the term behaves differently across markets — what we observe in Asia. We write definitions to be quotable in isolation, and we link to primary sources (platform documentation, research, and schema.org) rather than to other secondary glossaries.
Where the market data comes from
Quratic measures what AI search engines actually return to buyers in specific countries. Instead of calling a model's API from a single US datacenter, we open real browser sessions on residential IPs inside each target market — Singapore, Japan, Korea, Malaysia, Indonesia, and Hong Kong — across ChatGPT, Perplexity, Google AI Mode, Google AI Overviews, Gemini, Microsoft Copilot, and Claude.
This matters for a glossary because terms like AI Overview, LLM citation, and zero-click search do not behave uniformly worldwide. Rollout timing, language and script, and per-country model preference all change what a user sees. When an entry shows a figure, it reflects what we sampled in a named market over a stated period.
AI Visibility Score
The AI Visibility Score is Quratic's headline 0–100 metric for how present your brand is in AI-generated answers. It appears on the project dashboard, PDF reports, and email alert thresholds — alongside mention rate, which remains the underlying frequency measure in trend charts.
The score blends three inputs over your active prompts, models, and markets in the selected date range:
- Mention rate (55%) — the share of collected responses that name your brand. This is the dominant input; zero mentions yields a score of 0.
- Position (25%) — how prominently you appear when mentioned. Average list position is mapped to 0–100: position #1 scores 100; each step down reduces the position component by 12 points.
- Coverage (20%) — the share of distinct model × market combinations where you are mentioned at least once. Appearing only in ChatGPT Singapore scores lower than appearing across ChatGPT, Perplexity, and Google AI Mode in Singapore and Japan.
The weighted sum is rounded to the nearest whole number. Scores are computed per project and respect dashboard filters (date range, market, model, prompt category). Alert emails fire when the composite score drops below your configured threshold — not mention rate alone.
How data points are labelled
A figure only appears on a term page once it is verified against our collection. Each one states its market, sample, and time window. Terms for which we have not yet published a verified figure carry the definition only — we never estimate or invent a statistic to fill the slot.
Corrections and updates
Each entry shows its last-updated date. AI search is moving quickly, and we revise definitions and data as platforms change. If you believe a definition is wrong or out of date, email hello@quratic.com.