EntityMesh does not currently present a causal, cross-customer result showing that publishing a support hub increases AI citation rates. The implemented evidence is narrower: a diagnostic can record structural site signals, and separately configured monitoring can record responses for defined prompts and providers. A before/after comparison is meaningful only when the prompt cohort, provider, model, geography where available, timing, and capture method are compatible. Even then, an observed change is not automatically caused by the hub.
Structural readiness is not a citation outcome
Question-led answers, explicit entities, consistent facts, crawlability, and schema are inspectable publishing characteristics. EntityMesh can assess those characteristics. It cannot infer from them that a proprietary engine retrieved, trusted, or cited the page.
What EntityMesh can record
A completed diagnostic stores its own confidence and coverage limits. Configured LLM Presence or analyst-captured EchoScan evidence can record what a defined provider returned for a defined prompt. Public diagnostic research uses only unlinkable, exactly-once receipt-backed aggregate metric rows and describes that self-selected sample; it does not expose raw scans or treat the sample as a causal citation benchmark. No unpublished customer dataset or internal case study is presented as public proof.
What the evidence does not support is a guarantee
No tool can guarantee a citation-rate lift, because engines choose sources through systems outside EntityMesh. A causal claim would require a published methodology, compatible baselines, sufficient samples, controlled confounders, and reviewable evidence. Until that exists, the honest claim is structural readiness plus observed, scope-labelled monitoring—not citation lift.