Deepak Gupta markE-Books

Appendix B5 min readBy Deepak GuptaPublished

Glossary and Sources

Term Meaning in this book
AEO Answer engine optimization. Often used interchangeably with GEO; sometimes narrower, focused on direct answers.
Acting agent An AI agent that researches, compares, trials or buys on someone's behalf, rather than only answering a question.
Agent interface A machine-facing way for agents to query or use a product, such as an API or MCP server.
Authority layer Earned coverage, reviews, analyst and community mentions that engines use as third-party evidence.
Canonical description A fixed, approved description of the product at a set length, used verbatim across surfaces.
Capability exposure Making what a product can do legible to machines in public, structured, verifiable forms.
Citation rate How often one of your URLs is cited in answers to a prompt cluster.
Description accuracy Whether what an engine says about your product matches your truth sheet.
Description test The baseline audit step: asking engines to describe your product and scoring answers as missing, wrong, vague or accurate.
Entity The engine's accumulated picture of a company or product, built from many sources.
Exposure ladder Five rungs of capability exposure: hidden, described, specified, queryable, operable.
Fan-out An engine's practice of splitting one prompt into several related sub-queries before answering.
First-pick rate How often you are the first product an engine recommends.
GEO Generative engine optimization. The practice of improving how AI engines find, describe, cite and recommend you.
llms.txt A proposed Markdown file summarizing a site for language models. Widely adopted, rarely requested by AI systems in 2026.
MCP Model Context Protocol, an open standard for connecting AI agents to tools and data.
Mention rate How often your product is named in answers to a prompt cluster.
Naming map A public page linking each branded feature name to its standard functional term.
Product fact model The single owned source of canonical product facts, each with an owner, verification date and publishing locations.
Product layer The signals produced by the product itself: naming, docs, product data, self-description and exposure.
Product-resident GEO GEO focused on the product layer and owned by product and product marketing.
Prompt set The fixed list of buyer prompts you sample repeatedly to measure AI visibility.
Source mix The domains engines cite when they mention you, which shows where your entity is being built.
Truth sheet A signed-off one-page statement of what the product is, does and supports, used to score engine answers.
Vertical GEO GEO tuned to a domain's buyer language, authority sources and engines.

Sources

All sources were checked on November 6, 2026. Where a study was first reported by another publication, the link goes to the original.

Buyer behavior and traffic

How engines choose and cite

Volatility and measurement

Product layer: docs, data and exposure

Related resources from the author

About the author

Deepak Gupta founded LoginRadius in 2013 and scaled it to serve more than 1 billion users, helping define what is now called customer identity and access management (CIAM). In 2024 he started GrackerAI, which does Generative Engine Optimization for B2B cybersecurity vendors. He writes about AI, cybersecurity and B2B growth at guptadeepak.com.

Disclosure: GrackerAI sells GEO software and services. The criteria and methods in this book apply to it as they do to any vendor.