01 · OUTCOME & TRIGGER
Define what starts the work and what “done” means.
Separate the business outcome from the tool. Record the trigger, required inputs, completion definition, timing expectation and evidence that the outcome occurred.
Review the operating system before automating the mess.
A practical, evidence-first method for understanding how recurring work actually moves: what triggers it, who owns it, where decisions and handoffs occur, what happens on exception, what evidence proves completion, and which parts are ready for automation.
THE REVIEW SEQUENCE
Seven questions turn a vague workflow into something reviewable.
02 · CURRENT FLOW
Map the actual sequence and handoffs.
Trace the work across people and systems. Note queues, duplicate entry, waiting states, rework, manual transfers and places where context can disappear.
03 · OWNERSHIP & DECISIONS
Make responsibility and approval rights explicit.
Identify the outcome owner, supporting roles, approval points and decision rights. A process is fragile when responsibility changes without a visible handoff.
04 · EXCEPTIONS
Review the path when normal flow breaks.
List common failure modes, blocked states, unusual cases and escalation thresholds. Define who decides the next action rather than treating every exception as improvisation.
05 · CONTROLS & EVIDENCE
Ask what must be true before and after the action.
Distinguish preventive from detective controls. For each important control, identify the evidence a reviewer could use to confirm it happened or show that it remains unresolved.
06 · SIGNALS & REVIEW
Connect KPIs to decisions, not dashboards alone.
Choose the few measures that should trigger action, escalation or a recurring operating review. Record thresholds, owner, cadence and the decision each signal is meant to support.
07 · AUTOMATION READINESS
Automate stable rules; preserve human judgment.
Separate deterministic work from judgment-heavy decisions. Define inputs, outputs, permissions, failure handling, idempotency where relevant, human-review points and a safe fallback before automating.
OUTPUT
Sequence the smallest useful changes.
Turn findings into a short implementation backlog: control gaps first, avoidable handoffs second, automation candidates third. Assign owners and dependencies instead of producing a recommendation list with no operating path.
Artifacts that make the operating logic visible.
Current-state process & dependency map
Trigger, inputs, sequence, systems, handoffs, queues, dependencies and completion evidence.
Ownership & decision-rights matrix
Outcome owner, support roles, approvals, decision boundaries and escalation responsibility.
Exception & control view
Failure modes, thresholds, preventive or detective controls, evidence and unresolved gaps.
KPI & operating-review design
A small set of signals tied to a named owner, cadence, threshold and intended decision.
Automation-readiness backlog
Candidate tasks separated by rule stability, data availability, permissions, failure risk and required human review.
Implementation sequence
Prioritized actions with dependencies and ownership so recommendations can move into an operating cadence.
Evidence-first means no invented policy.
A process review should distinguish what is documented, what was observed, what is inferred, and what is still unknown. The method is designed to surface those gaps rather than fill them with assumptions.
A SMALL FICTIONAL EXAMPLE
An approval problem is rarely just an approval step.
Suppose a recurring onboarding workflow activates an account before a required approval is recorded. A process review would not simply recommend “automate approval.” It would first identify the trigger and owner, locate the approval decision, document the evidence that should exist, define the blocked-state exception and escalation path, then decide whether the approval can be enforced automatically or still requires human judgment.
This page describes an operating-review method and fictional examples. It is not an audit opinion, compliance certification, legal or accounting advice, a promise of specific results, or evidence of a client engagement. Technology and automation decisions still require testing in the actual environment.