NamedraGlossary
Plain definitions of the terms used in AI search visibility work. Each is written so it can be quoted on its own.
- AI search visibility
- How often and how favourably AI assistants (ChatGPT, Perplexity, Gemini, Claude, Google AI Overviews) name and cite a company when users ask for vendors in its category. Full explainer.
- Generative engine optimization (GEO)
- Increasing a brand's presence in AI-generated answers, mainly by earning citations on the sources engines retrieve from and by making the brand's own pages easy to quote.
- Answer engine optimization (AEO)
- Structuring content so answer engines can quote it directly: a direct answer in the first sentence, definitions, tables, FAQs, consistent facts and structured data.
- AI Visibility Score
- A 0-100 score computed as 100 × (0.40 × mention rate + 0.25 × share of voice + 0.20 × position score + 0.15 × citation rate). Formula and example.
- Mention rate
- The share of AI answers in which a brand is named by name, alias or domain.
- A brand's mentions divided by the mentions of all tracked brands in the same set of answers.
- Citation rate
- The share of AI answers that cite the brand's own domain as a source.
- Position score
- Credit for being named early: (K − position + 1) / K for K tracked brands; first mention scores 1.0.
- Grounding
- An AI model retrieving live web results before answering. Gemini calls it Google Search grounding; ChatGPT and Claude call it web search; Perplexity does it on every answer.
- Google AI Overviews
- The AI-generated summary with cited links at the top of some Google results pages. Shown for some queries only and varying by location and device.
- Entity consistency
- Using one name, one description and matching structured data (Organization schema, sameAs links) across the web so models connect every mention to the same company.
- Citation source
- A page an AI engine attaches to an answer as evidence. Review sites, community threads and comparison articles are the most common; vendor homepages are rare.
- Buyer-intent prompt
- A question a buyer asks an AI assistant while choosing a vendor: "best X for Y", "X vs Y", "Y alternatives", "how much does X cost". Library.
- Invisible prompt
- A prompt where competitors are named in the AI answers and the measured brand never is. The count of invisible prompts is the fastest summary of a visibility problem.
- llms.txt
- A plain-text file at a site's root that summarises the site and its key pages for AI crawlers and agents, like robots.txt for meaning rather than access.
- Run variance
- Differences between repeated answers to the same prompt, caused by sampling and by different pages being retrieved. The reason visibility is measured as a rate over several runs, not as a single answer.