Two years ago a chief executive looked at an organisation chart with seven layers and decided it should have four. The consultants redrew the boxes, the redundancy letters went out, and the saving showed up in the next set of results. It looked clean. The cost of that decision is arriving now, and it does not arrive as a line in the accounts. It arrives as a vacancy nobody can fill.
I wrote about the Great Flattening from the survivor’s side a few weeks ago, the view from the floor where the next round is aimed. This is the other side of the table. Once you have removed the middle of an organisation, something has to run what is left, and the people who ordered the cut are about to find out what they actually deleted.
The survivors are now carrying twelve people each
The average number of direct reports per manager rose from 10.9 in 2024 to 12.1 in 2025, according to Gallup’s analysis of span of control, close to a fifty per cent increase in team size since it first measured this in 2013. Gallup’s own work puts the point where engagement holds up at around eight or nine reports. A lot of managers are now well past that, doing the same coaching, the same one-to-ones, and the same problem-catching across half as much attention per person.
Korn Ferry’s 2025 survey found 41 per cent of employees said their employer had reduced management layers, and 37 per cent said the result left them feeling directionless. That second number is the one a board should worry about. Direction was part of the job you removed.
You did not remove the cost, you relocated it
The middle layer was doing work, even when it was hard to see on a productivity dashboard. It translated strategy into something a team could act on by Tuesday. It caught the small problem before it became a write-down. It carried the institutional memory of why the last reorganisation failed. Delete the box and that work does not vanish. It either stops happening or it falls on someone already at capacity.
Plenty of firms are discovering this in real time. Forrester’s 2026 Future of Work research found that 55 per cent of employers already regret AI-related layoffs, and the cautionary tales are piling up, with Klarna among the companies that cut hard, watched quality slip, and quietly rehired. The lesson is not that flattening is wrong. It is that flattening on the assumption that software already does the human half of management is a forecast, and a fairly optimistic one.
The succession bench you scrapped without noticing
Here is the part that surfaces eighteen months to two years after the cut, long after the consultants have invoiced and gone. A divisional VP resigns. You open the org chart to find the internal successor, and there is nobody close, because the roles that used to season people for that job are the roles you removed. Fortune put it bluntly: the middle-manager cuts saving you millions today are the ones that cost you the leadership pipeline by 2028.
The data backs the warning. Around 60 per cent of industrial firms report no ready-now pipeline for critical leadership roles. And the generation you would normally be grooming has been watching: only about 6 per cent of Gen Z say they want a senior leadership job, having spent their early careers seeing the management rung treated as the expendable one. You cannot flatten the development ladder and keep the leaders it used to produce. A board that has quietly done the first should be planning, today, for the second, whether through a deliberate board-readiness pipeline or external executive coaching for the people now expected to lead twice the span.
Redesign the structure, do not just subtract from it
Flattening done well starts with a question most cost-driven reorganisations skip: what was this layer actually doing? Map the real work before you delete the box. Where the work is routine coordination, software and a wider span can absorb it. Where the work is coaching, judgement, and developing the next cohort, a tight span earns its keep. The honest answer is usually that you should widen control in some functions and protect it in others, which is harder than drawing a single new number across the whole chart and calling it efficiency.
It also means investing in the people who survived the cut rather than assuming a wider remit is a promotion they should be grateful for. The capable operator now running fifteen people instead of seven needs decision rights that match the span, a clear remit, and a reason to stay. If you cannot give them that, the market will, and the executives most likely to leave a hollowed-out structure are precisely the ones you cannot afford to lose. For those weighing that decision from the inside, a clear-eyed executive career transition plan beats waiting to see whether the new structure was thought through. Some of them will find their next move on a board, which is where a sharp non-executive profile starts to matter.
Frequently asked questions
Is flattening the organisation always a mistake?
No. Many structures genuinely carry too many layers, and removing them speeds up decisions. The mistake is treating every layer as pure overhead and cutting by a blanket ratio, rather than mapping what each layer does and redesigning around it.
What is a healthy span of control?
It depends on the work. Gallup’s research points to engagement holding up at roughly eight or nine reports where managers are expected to coach and develop people. Routine, well-defined work can sustain a wider span. Twelve-plus reports with a heavy coaching expectation is where quality starts to slip.
How long before a flattened structure shows problems?
The financial benefit is immediate, which is why flattening is tempting. The structural cost typically surfaces eighteen months to two years later, usually as a senior vacancy with no internal successor and a thinner bench than anyone expected.
We have already flattened. What now?
Audit where spans have stretched too far, protect the roles that develop future leaders, and build a succession plan that does not rely on a pipeline you removed. A standing advisory relationship helps a leadership team think this through before the first unfillable vacancy forces the issue.
Flattening is not going away, and for many organisations it is overdue. What deserves more scepticism is the spreadsheet that treats a management layer as a cost with no output. The saving is real and it is visible this quarter. The cost is real too, and it turns up two years later wearing the face of a job you cannot fill from within. The leaders who come out of this well will be the ones who designed the new structure on purpose, decided which spans to widen and which to defend, and kept building the bench while everyone else was busy deleting it.
