Recommendability

Recommendability is the capacity of an entity, service, product, source, or organization to be recommended by an AI system under bounded, supportable, and contextually legitimate conditions.

This page is the canonical definition of Recommendability on Gautier Dorval. It is part of the phase 5 market bridge layer: a vocabulary layer designed to capture how teams, clients, dashboards, and AI-search tools speak before they reach the stricter doctrine of interpretive governance.

Short definition

Recommendability depends on stronger conditions than visibility or citability. A system must be able to identify the entity, understand its scope, compare it with alternatives without fabricating criteria, and avoid presenting a recommendation as if it were guaranteed, universal, or context-free.

The key point is that this term is useful only when it remains bounded. It names a real market-facing phenomenon, but it must not be treated as a guarantee of ranking, citation, recommendation, traffic, availability, or future system behavior.

What it is not

Recommendability is not a ranking position and not a brand preference. It is a response legitimacy threshold. A system can mention or cite an entity without having enough evidence to recommend it.

The distinction matters because AI-mediated search collapses several states that classical search kept separate: retrieval, citation, summary, comparison, recommendation, and decision support. A page can be retrieved without being cited, cited without being understood, understood without being recommended, and recommended without sufficient governing evidence.

Common failure modes

recommending an entity from weak or stale sources
comparing alternatives without declared criteria
turning a citation into an endorsement
recommending outside the stated service perimeter
failing to distinguish best-known, best-fit, and best-supported recommendations

These failures are not merely tactical SEO problems. They are representation problems. They show where a system may use a source, entity, or brand without preserving the conditions under which that use remains legitimate.

Why it matters

The term matters because answer engines increasingly move from information retrieval to decision support. Once a system recommends, it changes the risk profile of the answer and creates a stronger burden of evidence.

For market-facing search work, the term helps create an entry point. For governance work, it must be routed toward stricter concepts: canonical source, source hierarchy, proof of fidelity, interpretive observability, Q-Ledger, Q-Metrics, and answer legitimacy.

Governance implication

Recommendability should be governed by source hierarchy, response conditions, non-inference regimes, proof of fidelity, and explicit comparison criteria. The system must know when to recommend, qualify, abstain, or route the user toward a narrower question.

The practical implication is simple: do not let market labels govern the system. Use them to detect demand, observe symptoms, structure interventions, and route the work toward canon, evidence, auditability, source authority, and response conditions.

Phase 13 service bridge

This market-facing concept now has explicit service-market routes in the phase 13 layer. Start with AI visibility audits when the question is practical, commercial or diagnostic rather than purely definitional.

The phase 13 rule remains: a market label can capture demand, but it does not by itself prove visibility, citability, recommendability, answer legitimacy, service availability or correction success.

Phase 14 SERP ownership note

This page is the primary canonical definition target for Recommendability. Service, audit, glossary, framework, category, and article pages should link back here when they use this term.

Global routing: SERP ownership map.

Recommendability