DoneHat trust — clear claims and reviewable evidence
DoneHat reports sampled AI answer behavior. A visibility score is a diagnostic, not a promise that a model will recommend a product or that one content change caused a later answer.
Evidence before claims
Reports retain the checked buyer question, answer text when available, source receipts, failures, and confidence limits. Unsupported assertions are not treated as proof.
Human control
Agent-generated fixes are drafts for review. Customers decide what to publish and can track the implementation and later proof check.
Privacy and contact
See the privacy notice and terms for data practices, or contact us for a specific question.