Operational problems rarely announce themselves as one broken step. A handoff waits. A decision lives in one person’s memory. An exception returns to the inbox. Each feels manageable until the same work keeps coming back.
Before changing a process, choose one real item—a completed job, a customer request, a delivery, or an invoice—and follow it from its first signal to final close. Use the records people actually touch. Note who owns each step, where it moves, what triggers the next action, and where work pauses or information has to be reconstructed.
Ask a few plain questions: What was expected here? What actually happened? What evidence shows it? Who had authority to move it? What followed when it waited?
A short trace helps separate the symptom from possible causes. An overdue invoice may follow a missed handoff, but the underlying condition could be incomplete closeout, unclear ownership, or a status the next person cannot see. The trace does not prove which explanation is right. It gives you a better question to test against more examples.
Avoid prescribing automation before understanding the work. A new tool can make an unclear handoff faster to repeat. Sometimes the appropriate response is a smaller change, a clearer owner, or no build at all.
A useful operational review makes clear what is known, what remains uncertain, and what decision comes next. That keeps the response proportional to the evidence—and keeps people from carrying a process designed around assumptions.
Start with one item. Follow it honestly. Let evidence decide what deserves to change.
