Operational Complexity

Complexity grows because organisations prefer adding to changing. Adding a new procedure, system, or interface is easier than removing an old one. Stakeholders who benefit from the addition lobby for it. Stakeholders who would benefit from removal are diffuse and less vocal. The result is a system that grows thicker faster than it grows cleaner.

Mergers, regulatory changes, and technology transitions each leave sediment. Legacy processes coexist with new ones. Parallel systems run while migration is planned. Exceptions accumulate faster than standards. Each layer is defensible in its own terms. The cumulative effect is a system whose full state is unknown to any individual or team.

The people who operate critical systems rotate, retire, and move on. The system they inherit is usually more complex than the one their predecessors managed. Documentation drifts out of date. Training covers standard procedures, not edge cases. The ratio of system complexity to operator comprehension shifts steadily toward the system.

Comprehension is also fragmented. Different teams understand different parts. No one holds the entire operational model in working memory. Under normal conditions the seams between domains are invisible. Under stress they become the places where coordination fails and the system behaves in ways that surprise everyone involved.

Stress does not increase complexity; it reveals it. A system that operates smoothly in routine conditions may behave unpredictably when multiple exceptions coincide. The operators who would normally navigate one deviation find themselves managing several simultaneously, without a clear map of how they interact.

Incidents in complex systems tend to cascade in unexpected directions because the connections that matter are not the documented ones. A decision made to protect one subsystem may propagate harm to another in ways that were not anticipated because the interaction was not modelled. The system is not poorly designed; it has outgrown the understanding of its designers.

Reversibility is the capacity to undo a change and return to a known state. In a complex system this capacity degrades as interdependencies multiply. A change that once could be rolled back in hours now requires weeks of coordination across teams that no longer share a common understanding of the baseline.

The most resilient complex systems are those that invest in simplification as deliberately as others invest in expansion. They retire interfaces, consolidate exceptions, and periodically reconstruct operational knowledge from first principles. This work is expensive and politically difficult, but it is the only counterweight to the natural drift toward complexity.