Glossary.AI search glossary.
Short definitions of the terms used in AI search and in Benchmarq reports, from GEO, AEO, and AIO to llms.txt, AI crawlers, and MCP. Each term has its own link, so you can share it.
What do the terms mean?
In alphabetical order. Track names follow Benchmarq: GEO is Generative Engine Optimization, AIO is AI Index Optimization, AEO is Answer Engine Optimization, and CORE is Technical health.
A
- ACP (Agentic Commerce Protocol)
- An open protocol that lets an AI agent place an order with a store on a user's behalf. Benchmarq checks for ACP support from the DAE 1.4.1 audit set onwards. E-commerce suite
- AEO (Answer Engine Optimization)
- Making your content easy for answer engines and AI agents to read and quote: direct answers, clear structure, and semantic clarity. AEO is one of the four Benchmarq tracks. What is AEO?
- Agentic App (ARO)
- A Benchmarq add-on that tests whether AI agents can log in to your site or app and take action, such as filling in a form or completing a task. It is reported separately and does not change the overall score. Agentic apps suite
- Agentic Commerce (ACO)
- A Benchmarq add-on that tests whether an AI agent can complete a purchase on your site: product data, checkout, and payment protocols such as AP2. It is reported separately and does not change the overall score. E-commerce suite
- AI agent
- Software that uses an AI model to pursue a goal on a user's behalf, for example by browsing pages, reading data, filling in forms, or buying something. An agent needs a site it can read and use.
- AI crawler
- A bot that fetches web pages for an AI product, either to train a model, to build an index, or to answer a question. Examples of crawler user agents are GPTBot (OpenAI), ClaudeBot (Anthropic), and PerplexityBot. You control their access with robots.txt. AI crawler access
- AI Overviews
- Google's AI-generated summaries at the top of some search results, with links to the sources they used. Not to be confused with AIO, which is how Benchmarq abbreviates AI Index Optimization.
- AI readiness score (overall score)
- The single 0 to 100 score in a Benchmarq report. It combines the four track scores (GEO, AIO, AEO, and CORE). Add-on results are reported separately and are not part of it. Benchmarq reports
- AIO (AI Index Optimization)
- How discoverable your pages are in AI-powered search results: whether AI indexes and crawlers can reach, index, and place them. AIO is one of the four Benchmarq tracks. Some sources also use AIO for Google's AI Overviews. What is AIO?
C
- Canonical URL
- The preferred address of a page, declared with a link element or an HTTP Link header. It tells crawlers which of several similar URLs to index and cite.
- Citation
- A link or reference to your page in an AI-generated answer. Whether an engine cites a site is outside your control; you can only make the site easier to find, read, and quote.
- CORE (Technical health)
- The foundation track in Benchmarq: crawlability, structured data, and parser performance. If crawlers cannot fetch and parse a page, the other tracks cannot help. CORE checks
D
- DAE (Discovery & Authority Engine)
- The audit framework behind Benchmarq, built by Enso DX: an API with standardized tests that scans a site and scores it. Each audit set has a version number, such as DAE 1.4.1. Audit engine
- Deterministic test
- A test that runs fixed rules and gives the same result for the same input. The core Benchmarq tests are deterministic, so a score change means the site changed. How the score is built
G
- GEO (Generative Engine Optimization)
- Improving how well AI models cite and surface your brand in generated answers. GEO is one of the four Benchmarq tracks. What is GEO?
J
- JSON-LD
- A way to write structured data as JSON inside a script tag, using the schema.org vocabulary. It is the format search engines and AI systems most often read for structured data. Schema for AI search
L
- llms-full.txt
- A proposed companion to llms.txt that holds the full text of your key pages in one file, so a language model can read them without crawling each page. llms.txt explained
- llms.txt
- A proposed plain-text file at the root of a site that summarizes what the site is and lists its most useful pages for language models. It is a convention, not a standard that every engine follows. llms.txt explained
M
- MCP (Model Context Protocol)
- An open standard that lets AI assistants call tools and read data from other services. The Benchmarq MCP server lets an assistant read your scores and start page scans. Benchmarq MCP
P
- Prompt tracker
- A tool that runs a set of prompts against AI engines and reports whether and how your brand appears in the answers. It shows what AI says, not why your site is or is not used. Product comparisons
R
- robots.txt
- A text file at the root of a site that tells crawlers which paths they may fetch. Each crawler is addressed by its user agent, so you can allow some AI crawlers and block others. AI crawler access
S
- Schema.org (structured data)
- A shared vocabulary for describing what a page is about, such as an organization, a product, a FAQ, or an article, in a way machines can read. It is usually added as JSON-LD. Schema for AI search
T
- Track
- One of the four scored areas of a Benchmarq audit: GEO, AIO, AEO, and CORE. Each track has its own score from 0 to 100. Audit engine
U
- UCP (Universal Commerce Protocol)
- An open standard for agentic commerce that lets AI agents find products and complete a checkout with a store. Benchmarq checks for UCP support from the DAE 1.4.1 audit set onwards. E-commerce suite
W
X
See these terms in your own scores
Run a Benchmarq scan and get a score for each track, with the fix behind every deduction.