Framework

The Answerability framework.

The definitions the rest of the practice is built on — what we measure, and why a citation depends on all three.

Answerability is the composite a company earns when an AI system decides whether to name it: the ability to be retrieved, trusted, and used as an answer to a buyer's question. It is earned across three pillars — Content, Retrieval, and Trust — and it is constrained by the weakest of them, not their average. A page can clear two pillars and still never be cited because it fails the third. We do not abbreviate the framework to an acronym; we spell out the pillars.

Pillar 01

Content

Is there content that answers what buyers actually ask, in a form an engine can lift — coverage plus answer-shape?

Pillar 02

Retrieval

Can the engines access, crawl, parse, and structurally understand that content in the first place?

Pillar 03

Trust

When several sources could answer, do the engines treat yours as cite-worthy — internal evidence plus external corroboration?

The canon

The working primer defines the category; each pillar note is the long-form, dated reference for one axis, with real cross-engine captures. These are written to be quoted — by readers and by machines.

How we apply it

The framework is the lens; two companion pages show it at work. The methodology is how the pillars are measured against a standing prompt set across five engines; the glossary fixes the vocabulary the practice uses consistently.

From definitions to evidence. The framework says what to look for; the Answerability Index shows it in real cross-engine data, sector by sector. When you want it run on your own company, that's the Diagnostic.

Last updated · hello@answerability.ai