Pillar · 1
Content
Whether a company has content that answers what buyers actually ask, in a form an AI engine can lift as a self-contained answer. Coverage plus answer-shape.
Glossary
The vocabulary of machine-mediated discovery.
The terms we use to describe how AI systems retrieve, trust, and cite companies — defined plainly, so they can be quoted, argued with, and reused.
Signature concept
Retrieval Surface
The specific slice of documents, pages, and entity records that AI systems can reach, parse, and repeatedly surface when answering questions about a company — the machine-readable shadow your brand casts across embeddings indexes, structured feeds, and agentic search layers. It is not the same as "your website."
Composite
Answerability
A company's ability to be retrieved, trusted, and used as an answer by AI systems — the composite measure at the center of our framework. It is constrained by its weakest pillar (Content, Retrieval, or Trust), not the average of the three: a citation requires all three to hold.
Pillar · 2
Retrieval
Whether AI systems can access, crawl, parse, and structurally understand your content. Distinct from your Retrieval Surface, which is the resulting reachable set across the wider machine layer. (Always "Retrieval," never "Retrievability.")
Pillar · 3
Trust
Whether AI systems treat a source as cite-worthy — a function of internal evidence (named authorship, stated methodology, inline citations) and external corroboration (resolvable entities, third-party coverage). The pillar that most often decides the close calls, and the slowest to build.
Concept
The plural answer layer
AI-mediated discovery is not one system but many. The answer layer is plural, embeddings-driven, and agent-consumed: different engines query different indexes through different retrieval paths, so the same question can resolve to different answers — and a company can be visible in one and absent from another.
Concept
Query fan-out
The practice by which an AI system rewrites a single user prompt into multiple targeted search queries, runs them, and merges the results into one answer. The practical consequence: a company is evaluated across a tree of sub-queries it never sees, not against the one question the user typed.