What Is an Entity, and Why AI Engines Need One Before They Trust You
An entity is a resolved, machine-readable identity distinct from your website. Without one, engines read every mention of your name as an unlinked string.
AEO Fundamentals
The definitions, the mechanics, and the vocabulary. Start here if AEO is new to you, or if you have been doing it for a year and want to check that the terms you are using mean what you think.
This section covers what an AI citation actually is and why it behaves nothing like a backlink, how each of the ten major engines retrieves differently, and the difference between being mentioned, being cited, and being recommended. Those three get used interchangeably by most tools and they have three different fixes.
Everything here is written to be checked. Where a claim comes from a study, the study is linked.
16 posts
Mention Rate vs Citation Rate vs Recommendation Rate
Three different outcomes with three different fixes. A mention can come from training memory, a citation needs a live fetch, and only one of them sells.
What Is Share of Voice in AI Search, and How to Calculate It
AI share of voice is the share of answers in a fixed prompt set naming your brand, split across named competitors. The method, and what moves it wrongly.
How Copilot Inherits Bing’s Index, and What That Means for You
Bing Webmaster Tools reports Copilot citations, the pages cited, and the queries used to retrieve them. It is the only free first-party AI diagnostic.
How Gemini Grounds Answers in Google Search
Gemini grounds in Google’s index, so Search eligibility is the entry ticket. That direct relationship is unusual, and a bad proxy for the other engines.
How Claude Retrieves and Cites Sources
Claude searches only when a question needs current information, and every citation carries the exact passage used, not just the page it came from.
Google AI Overviews vs AI Mode: Two Systems, Two Sets of Rules
AI Overviews largely summarises what already ranks. AI Mode fans the question out into related searches first, so it names brands Overviews never will.
How Perplexity Ranks Sources Differently From Every Other Engine
Perplexity shows its sources inline, which makes it the best engine to audit your category with and the worst one to generalise a citation rate from.
How ChatGPT Chooses Sources: Browsing, Memory, and Training Data
ChatGPT answers from three separate stores that update on three different clocks, and the phrasing of the question decides which one you get.
The AI Engines Your Buyers Actually Use, and How Each One Retrieves
Three retrieval families sit behind every AI answer: live web search, grounded index, and training-data recall. Each one fails you differently.
What Is Retrieval Augmented Generation and Why It Decides Your Visibility
RAG means the model fetches documents before answering. Visibility splits in two: whether the retriever finds you, and whether the generator names you.
How Large Language Models Decide Which Brands to Recommend
Recommendation happens in two stages. Retrieval decides whether you are found, and generation decides whether you are named. Consistency beats prominence.