Published
Updated
How this reference is organised
This site has three kinds of page and one machine-readable file, and the whole of it fits on this page. There are no tags, no archives, no pagination and no dead ends: every page is reachable from the navigation in at most two clicks, and every page is listed below.
Entity pages are the reference layer. Each one covers a single subject — one AI search surface, or one technique that publishers are told to adopt — and each leads with an evidence verdict before any prose. They sit at the root of the site, one directory deep, because they are the pages the site exists to publish and nothing should stand between them and a reader arriving from a search result. They are split across two category indexes, and a single directory lists all of them together.
Guides are the explanatory layer. They answer a question that spans several entity pages — how engines choose what to cite, what changed when AI Overviews launched, how to build a visibility report — and they link down into the reference pages that carry the detail. All twelve live at a flat path under the guides directory.
Pillar guides are the three hubs that hold the guides together. Each one covers its silo end to end and links to the four guides underneath it; each guide links back up to its pillar and sideways to its three siblings; and the breadcrumb on every guide names its pillar. The silo is expressed entirely in the link graph rather than in the URL, which keeps every guide one directory deep and keeps a guide's address stable if it is ever moved between silos.
Why the URLs look like this
Every page on this site is a directory with a trailing slash and no file extension. Every internal link and every asset reference is root-relative, beginning with a slash, so a page resolves identically no matter what depth it is served from and no link on the site depends on where the reader currently is. Nothing on the site links to a relative parent path.
The two category indexes and the directory are the only listing pages. There is no region scheme, no tag cloud and no date archive, because none of those would carve this subject along a line a reader thinks in. The two axes that matter here are what a page is about — a product, or something a publisher does — and how much evidence stands behind it, and both of those are visible on every index and on every card.
Dates, and what they mean
Every page carries a publication date and a modification date, and the XML sitemap reports each page's real modification date rather than the date the site was last built. That distinction matters: a build changes a stylesheet fingerprint and a footer year on every page at once, and if the sitemap reported that as a change to fifty-one documents, the signal would be worthless. The modification date on a page moves when the words on the page change, and at no other time.
The machine-readable layer
Alongside the pages, this site publishes an entity map: a structured description of the subjects covered here, the relationships between them, and short extractive passages drawn from the pages themselves, each attributed to the page it came from. It is linked from the head of every page, from the footer, and from the robots file, and it is readable both as JSON and as an ordinary web page.
The entity map exists because a reference site's value to a retrieval system is in how cleanly its subjects and their boundaries can be read. Every passage in it is extracted from a real page rather than written for the file, every relationship points at an entity the file itself defines, and every external identifier attached to an entity was verified rather than guessed. Where no verified identifier exists, the entity carries none.
Everything on this site
Main pages
AI search surfaces (7)
Techniques and signals (17)
- Generative Engine Optimization
- Answer Engine Optimization
- GEO vs AEO vs SEO
- AI Citations and Source Selection
- Brand Mentions in AI Search
- Schema Markup for AI Search
- FAQ Markup and Answer Extraction
- Content Chunking and Passage Retrieval
- Entity Grounding and the Knowledge Graph
- E-E-A-T Signals in AI Search
- AI Crawler Directives and llms.txt
- Retrieval-Augmented Generation
- Query Fan-Out
- AI Visibility Tracking
- AI Overview Rank Tracking
- Share of Voice in AI Search
- AI Search Attribution and Analytics
Understanding AI Search
- Understanding AI search (pillar guide)
- How AI search engines choose what to cite
- What changed when AI Overviews launched
- Why AI search traffic behaves differently
Optimizing for AI Search
- Optimizing for AI search (pillar guide)
- How to Rank in AI Overviews
- Structuring Content for Answer Extraction
- Technical Requirements for AI Crawlers
- AI Search Visibility for Law Firms
Measuring AI Search Visibility
- Measuring AI search visibility (pillar guide)
- How to track AI Overview visibility
- How to evaluate an AI visibility tracking tool
- What AI search analytics can and cannot tell you
- Building an AI visibility report that survives scrutiny