The words, in plain language

Definitions

AI visibility

How often, and how prominently, a business is named in answers generated by AI systems such as ChatGPT, Gemini, Perplexity and Google AI Overviews, as distinct from how it ranks in a list of blue links.

Share of Model

The percentage of a defined set of relevant queries in which a property is named, cited or recommended in AI-generated answers. Measured per property and per engine, against a named set of five competitors, on a query set of at least 25 queries fixed for the engagement, and reported monthly.

CITED

drishva.ai's five-component framework for AI visibility work: Crawlability, Identity, Trust, Expression and Distribution. Every delivery hour is assigned to one component, which makes the work estimable and auditable.

Generative Engine Optimisation (GEO)

The practice of improving how often and how accurately a brand appears in AI-generated answers. Distinct from search engine optimisation, which targets ranked lists of links.

Answer Engine Optimisation (AEO)

Used interchangeably with GEO by most practitioners. Where a distinction is drawn, AEO refers to direct-answer surfaces and GEO to generative synthesis.

Citation

A source an AI system draws on and, usually, links to when producing an answer. Citations are the mechanism by which a property enters an answer, and most of them are not the property's own website.

Earned media

Coverage a brand did not pay for or publish itself: journalism, guides, tourism board listings, review platforms, forums. Roughly 84% of AI citations come from earned media rather than brand-owned content.

Entity

A machine-readable record of a real thing — a hotel, a company, a place — that AI systems and search engines can resolve unambiguously. A property with a fragmented entity signal can be treated as several unrelated businesses.

Entity resolution

The process by which a machine decides that two mentions refer to the same real-world thing. Where a group operates under several brand names, resolution frequently fails, and the group's properties are treated as independents.

Crawlability

Whether an AI system is permitted and able to read a website's content. Commonly blocked by content-delivery-network bot rules that override the site's own robots.txt.

Structured data

Machine-readable markup, usually schema.org JSON-LD, that states explicitly what a page is about rather than leaving a machine to infer it.

llms.txt

A plain-text file at the root of a domain that summarises a site for AI agents, in the way robots.txt states crawl permissions. An emerging convention rather than a standard.

Grounding

The retrieval step in which an AI system fetches live sources before generating an answer, rather than answering from training data alone. Grounded answers carry citations.

Query set

The fixed list of questions used to measure Share of Model. Locking it for the duration of an engagement is what makes month-to-month comparison valid.

Competitive set

The five named properties a subject property is measured against. Chosen at baseline with the client and held constant.

Citation decay

The tendency of an AI system's cited sources to change over time. Independent measurement has found roughly 65% of cited sources change from one day to the next, which is why visibility is held rather than built once.

Source market

The country a traveller is searching from. AI answers diverge substantially by location, so a query run from the wrong country produces a misleading measurement.

Visibility Snapshot

drishva.ai's free assessment: five questions a guest would realistically ask about a market, run across four AI engines, returned with screenshots and the sources cited.

If we have used a term on this site that is not defined here, tell us and we will add it. Definitions are ours unless attributed; where a figure appears, its source is given on the page it appears on.