Large organisations restructure repeatedly, often reversing an arrangement they adopted a few years earlier. The cycle is a consequence of the trade-offs built into any structure.
Every structure sacrifices something
Organising by function groups specialists together, which builds deep expertise and consistent standards while making cross-functional work slow and political.
Organising by product or region gives each unit end-to-end control and speed, at the cost of duplicated capability and divergent practices across the company.
Neither is wrong. Each solves the problem the other creates, which is why organisations oscillate between them rather than converging on an answer.
The neglected cost accumulates until it dominates
A newly functional organisation enjoys consistency, and over time the complaints shift to slowness and internal handoffs.
A newly divisional organisation enjoys speed, and over time the complaints shift to duplicated teams, incompatible systems and inconsistent customer experience.
By the time the complaints are loud enough to force change, the alternative structure's costs have been forgotten, which makes the switch feel obviously correct.
New leaders arrive with a mandate to act
An incoming executive is expected to demonstrate change, and restructuring is fast, visible and entirely within their authority.
It also allows a leader to place trusted people in key positions and to remove layers without a public performance dispute.
Because the effects take years to evaluate and the leader is judged sooner, the incentive favours reorganising over persisting with an inherited design.
Matrix arrangements defer the choice
Reporting into both a function and a business unit is an attempt to keep the benefits of each, and it works where priorities between the two are genuinely aligned.
Where they conflict, the employee is left to resolve a disagreement between two managers, and decisions escalate rather than being made.
That pushes organisations back towards a clearer primary axis, which restarts the cycle from a different point.
The transition itself has a real cost
Productivity falls during a reorganisation as reporting lines change, priorities are renegotiated and people spend time securing their positions.
Institutional knowledge leaves with the people who depart, and relationships that made informal coordination work are broken and rebuilt slowly.
Which means a structure only pays for itself if it lasts long enough to recover the disruption, and frequent restructuring can lose money even when each individual change is well reasoned.