Record current effort
Capture process frequency, people involved, manual steps, review time, exceptions, rework, and the source for each observation.
This page focuses on evidence-backed assumptions, control dependencies, and review notes. It does not present public price ranges or sales math.
This route is for evidence planning. A value claim is only useful when the source, assumption, confidence level, and control dependency are visible.
Capture process frequency, people involved, manual steps, review time, exceptions, rework, and the source for each observation.
Explain which step changes, which system or agent assists it, and what remains explicitly human-owned.
Separate directly observed facts from interview estimates, vendor assumptions, pilot findings, and unsupported aspirations.
Document any approval, access, logging, disclosure, rollback, or monitoring control required before the value claim can be trusted.
Value claims should have a business owner who understands the baseline and can challenge the assumption.
Evidence decays. Record when the claim should be checked again and what signal would prove or disprove it.
Use stakeholder interviews and evidence requests to ground the value case in observable process facts.
WorkflowMap current work so the value claim points to a real process, not a generic productivity assumption.
ControlShow which controls are strengthened or weakened by the proposed AI-assisted change.
DisclosureUse disclosure guidance when the value claim depends on a public explanation of AI use, human review, or data handling.
BenchmarkCompare agent deployment readiness when the proposed value comes from connected or action-taking systems.
ProofCheck whether the final claim has source, scope, owner, control, date, and public-answer support.
Tie the claim to the baseline, the owner, the control dependency, and the review date before it becomes part of a public record or deployment decision.