AEO Glossary
Concise definitions for the core concepts behind evidence-gated answer-engine diagnostics. Quotable, neutral, and intentionally restrained.
Answer readiness
The condition in which a page presents a clear, self-contained answer that an answer engine can lift without reconstructing meaning from scattered prose.
What this meansThe first viewport of the page resolves the user's question in plain language — not after several paragraphs of brand copy.
Common misunderstandingAnswer readiness is not about FAQ blocks. A page can be FAQ-heavy and still unready if no passage stands alone as an answer.
Eligibility
Whether a page meets the structural conditions required to be considered as a source by an answer engine.
What this meansEligibility is about being in the candidate pool — not about being chosen. A page can be eligible and still not selected.
LimitationEligibility scores describe structural readiness only. They do not predict ranking or citation outcomes.
Representation drift
Inconsistency between how a page describes an entity and how external systems represent the same entity.
What this meansThe business name, service area, or category drifts between the page, schema, knowledge graph, and third-party listings.
Common misunderstandingDrift is not always a content problem. It is often a coordination problem across surfaces.
Entity clarity
The unambiguity of the primary subject of a page across visible content, structured data, and internal references.
What this meansA reader and a parser arrive at the same answer to "what is this page about?"
Common misunderstandingEntity clarity is not solved by adding more keywords. It is solved by stable, consistent reference.
Evidence-gated diagnostic
A diagnostic system that refuses to issue recommendations beyond what observed evidence supports, and labels confidence on every claim.
What this meansIf the page or available data does not support a recommendation, the diagnostic withholds it rather than guessing.
Common misunderstandingEvidence-gated does not mean conservative. It means honest about what is known versus assumed.
Recommendation confidence
A label attached to each recommendation indicating whether it is observed from the page, inferred from indirect signals, or unsupported by sufficient evidence.
What this meansRecommendations carry one of: Fix Now, Improve, Well Covered, or Verify before action.
LimitationConfidence reflects the strength of the supporting evidence — not the size of the expected impact.
Page identity
The stable, declared subject of a page — the entity, service, or topic the page is about — preserved across edits and recommendations.
What this meansIdentity is locked before recommendations are generated, so the diagnostic does not quietly reframe the page into something it is not.
Common misunderstandingIdentity preservation is not anti-change. It is anti-drift.
Answer-engine visibility
The extent to which a page is selected, cited, or used as a source by AI-driven answer systems.
What this meansVisibility here is not impressions or rank — it is whether the page contributes to the generated answer.
LimitationVisibility is observable in some surfaces (Google AI Overviews, citation panels) and invisible in others. Tracking is partial by nature.
Trust signals
Verifiable on-page evidence — authorship, dates, sources, certifications, service area — that supports the page's claims.
What this meansA trust signal is only useful if it is verifiable. Unverifiable claims are weaker than no claim at all.
Common misunderstandingTrust signals are not badges. A logo is not evidence; a verifiable certification number is.
Observed vs inferred
A distinction between facts measured directly from the page or from search data and facts deduced from indirect signals.
What this meansObserved claims are reproducible. Inferred claims are plausible interpretations and are labeled as such.
LimitationInferred findings are useful as hypotheses, not as conclusions.
Confidence-aware recommendation
A recommendation that exposes its supporting evidence and labels itself Fix Now, Improve, Well Covered, or Verify before action.
What this meansThe team reviewing the recommendation can see why it was issued and how strongly it should be acted on.
Common misunderstanding"Verify before action" is not a soft fix-now. It is a refusal to recommend until missing facts are confirmed.
← AEO Hub
·
Compare approaches
·
Home