Where the designed process doesn’t cover reality. Mapping where they cluster is one of the most direct ways to find root cause — the set points at the same place in the structure.
Is this the same centrality logic as the graph?
Yes. When different parallel controls, created by people who never spoke to each other, point at the same system or the same step, that concentration is the signal — not the volume of complaints about each one in isolation. See how this works in why the layer is a graph.
Does a parallel control point at a person or at a process?
At a process, almost always. The person who created the control merely solved, at their own end, what the structure left unsolved — and the finding is investigated on the process, never on whoever made the adjustment.
Is this an audit?
No. An audit checks compliance against a rule and produces a verdict. Mapping parallel controls shows where practice diverges from the design, as a question to investigate — without concluding that anyone was at fault.
What do you do with the map of parallel controls?
Decide, together with whoever owns the area, whether the exception should become official process, an automation, or whether the informal control already is the right answer and only needed to be acknowledged — what the page what else happens calls a practice nobody had formalised.
A parallel control rarely appears alone. It has relatives — other exceptions, in other areas, solving variations of the same problem without knowing the others exist.