Small-business outsourcing research

Auditing process-map drift after a workflow change

A method for checking whether a small business process map still matches the system, owners, and exceptions used in daily work.

## Question and scope This audit asks whether a process map still reflects the current workflow after a system or ownership change. It examines the map, the source systems, and a sample of real cases. It does not evaluate the person who drew the map. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Question and scope Set the scope to one process and one change event, such as a new tool, a new approval threshold, or a team reorganization. Record the map version and the effective date of the change. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Population and unit The unit is one process step or handoff in the map. A map with ten steps produces ten units. Record the step, the owner, the system, the input, the output, and the exception path. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Population and unit Include steps that were removed or added after the change. Exclude unrelated processes, with a count. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Definitions and measures Define drift as a difference between the map and the work observed in the sample. Record missing steps, wrong owners, outdated system names, and missing exception paths. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Definitions and measures Separate a documentation error from a process error. A map can be wrong while the work is sound; a map can be accurate while the process is inefficient. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Evidence standard Collect the map, the current system configuration, the role list, and three recent cases. Compare the map with the case trail. Record the evidence type and the date. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Evidence standard Mask customer and employee details in the analysis table. The study needs the step and the owner, not the identity of the person in the case. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Sampling and collection Review every step in the selected process, then test the map against the cases. If the process is large, sample steps by risk. Use a documented selection rule. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Sampling and collection Include an exception case and a case that crossed a shift. Those cases expose handoff and ownership drift that a routine case may hide. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Analysis and reporting Report the number of steps, the number with drift, the number with a missing owner, and the number with an outdated system. Show the results by step type. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Analysis and reporting Inspect the drift cases. Was the change not communicated, the map not owned, or the system name copied from an old guide? The pattern should lead to one update or control change. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Limitations The map may intentionally simplify a complex process. A difference is not always an error. The reviewer should distinguish a deliberate summary from a missing step that changes the work. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Limitations The case sample may not cover every branch. State the coverage and avoid claiming the map is complete when only the common path was reviewed. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Review and next action Choose one change: a map owner, a review trigger for system changes, a required exception branch, or a quarterly walkthrough. Assign the owner and the next review date. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

## Review and next action Update the map and test it on the next case. Close the review when the owner accepts the result and the next check is scheduled. Record the evidence and the next owner before the item leaves the queue. Keep the note factual enough for a later reviewer to continue.

Sources

Frequently asked questions

Does a favorable sample prove the routine caused the result?

No. The design is observational and supports a bounded operating decision, not a universal causal claim.

What should an owner review first?

Start with high-consequence defects, missing denominators, exceptions, and cases that required the owner to reconstruct the evidence.

Related research