Cost discipline and efficiency efforts have recently led many global companies to reduce organizational layers and flatten structures. Often described under labels such as “unbossing,” the approach promises lower overhead, shorter reporting lines and faster decisions on paper.
But reducing layers is not the same as reducing operational leadership. When middle-management roles are treated merely as approval points or meeting coordinators, short-term cost gains can create longer-term pressure on decision quality, coordination and institutional memory.
The balance-sheet illusion and hidden costs
From a CFO perspective, middle-management layers may appear as overhead with relatively high compensation and no direct revenue generation. Reducing positions can therefore lower personnel and management costs in the short term.
Yet the calculation does not directly show weaker decision quality, operational friction or the time senior leaders spend on micro-decisions. One core role of middle managers is to translate strategic objectives into executable work packages, connect teams, resolve problems before they escalate and carry weak signals from the front line upward. When that buffering mechanism weakens, small operational bottlenecks begin to move up the hierarchy.
Risk of decision fatigue at the top
When layers are removed, managers may inherit wider spans of control. The critical point is not to search for a single ideal number. As McKinsey’s organizational design work also emphasizes, the right span depends on the role, the complexity of the work and how independently the team can operate. The problem begins when more people are simply assigned to the same manager without redesigning the work.
If a senior executive who should focus on strategy, capital allocation and macro risks increasingly has to deal with delivery schedules, team priority conflicts or performance issues, the organization chart may become thinner while the decision queue gets thicker. A company may believe it has accelerated while more decisions accumulate in the calendars of a small number of senior leaders, creating a new bottleneck.
Management literature has long discussed that organizational flattening does not automatically push decisions downward and, in some structures, may instead concentrate decisions at the top.
The quiet drain of institutional memory
Layoff programs do not remove only positions. They may also remove people who know who works well with whom in a crisis, which supplier has shown flexibility, where teams have encountered bottlenecks and how work actually gets done behind formal procedures.
Middle managers often carry a significant share of this informal nervous system. When this layer is emptied quickly, mentoring and context for new employees can weaken, while junior teams may face greater uncertainty when taking initiative. The issue is therefore not only “How many managers do we need?” but “Which management capacity are we losing?”
The relational gap AI cannot fill
In delayering programs, generative AI is increasingly expected to absorb coordination, reporting and information-summarization work. AI can summarize budget variances, extract meeting notes, track workflows and speed up access to information.
But management is not only information processing. Interpreting the team dynamics behind a delay, quiet resistance on the customer side, loss of trust or conflicts of interest within the organization requires relationships, experience and judgment as much as data. Today’s AI systems can support that relational context, but they do not own it. If experienced management judgment is removed completely, companies may end up with more data without expanding decision capacity to the same degree.
A sustainable organizational model
Every company can accumulate unnecessary approval chains and hierarchical residue over time. Delayering is therefore not inherently wrong. The risk is using the savings target as a substitute for organizational design.
- Automate administrative approvals through systems and workflow architecture where possible, while preserving leadership roles that manage people and projects.
- Before widening spans of control, assess role complexity, team maturity and whether decision authority has genuinely been delegated downward.
- Evaluate savings not only through compensation cost, but also through longer decision cycles, rework, employee engagement and senior-management time spent on micromanagement.
- Reframe middle management from a control layer into a leadership layer focused on coaching, prioritization, problem solving and cross-team coordination.
Simplifying hierarchy is not the same as weakening the organization’s backbone. Companies that eliminate critical management capacity for short-term cost improvement may later repay the same cost through delayed decisions, rework and loss of strategic focus. The objective should not be fewer managers at any cost, but stronger management capacity with less bureaucracy.
Research note
The editorial review removed weakly generalizable numerical claims such as a fixed “ideal span of control” or universal adaptation periods. The studies below discuss why management spans should be designed around role and work complexity, how delayering can shift decisions upward in some structures, and why strong middle managers matter to organizational health.