Three years ago, a CEO I ghostwrote for called me the night before announcing a restructuring that would cut forty percent of her management layer. She wasn’t nervous about the press release. She was nervous about who would run the company in a decade, because she had just realized nobody below her had done the job that took her fifteen years to learn.
I’ve thought about that call more than almost any other client conversation I’ve had. It wasn’t about the org chart at all. It was about a leader recognizing, in real time, that she was about to remove the exact structure that had made her.
That is the leadership pipeline crisis, and it is not something happening to companies from the outside. It is something executives are building themselves, often for good reasons, without asking what the cut removes.
Why the Leadership Pipeline Crisis Is a Mirror, Not a Memo
Gartner projects that by the end of 2026, one in five organizations will have eliminated more than half of their middle management roles. Oracle alone has cut between 20,000 and 30,000 positions, most of them managers. Meta and Google have made comparable moves, and the trend shows no sign of slowing.
These are not struggling companies making desperate cuts. They are some of the most profitable organizations in the world, choosing to remove a layer of leadership development on purpose, to fund AI investment they consider more urgent.
That distinction matters, because it means the leadership pipeline crisis is not an accident of bad economics. It is a deliberate trade-off, made by people who assumed the pipeline would take care of itself.
The logic sounds clean on a slide. Fewer layers, faster decisions, lower cost. For a quarter or two, the numbers back it up, so the decision looks obviously correct.
But I’ve sat across from enough executives making this call to notice something the spreadsheet misses. Every one of them built their own judgment through the exact roles they are now removing. They learned to run a hard conversation, defend a budget under pressure, or make a fair call on letting someone go, because a manager three levels down was once forced to do those things badly, then better, then well.
Cut that layer and you are not simply flattening a structure. You are removing the place where your successor was supposed to be built. That is the belief this piece needs to disturb: efficiency and leadership development are not the same project, even when they share a budget line.
What Happens When Companies Eliminate Middle Management?
Companies that eliminate middle management lose far more than headcount. They lose the informal coaching, context-setting, and career guidance that used to live inside those roles, and employees notice the gap almost at once, long before it shows up in any retention report or engagement score.
Korn Ferry’s 2025 Workforce Survey found that 41 percent of employees say their company has reduced management layers. Of those, 37 percent now feel directionless, because the people who once translated strategy into daily decisions are simply gone.
That statistic matters because it shows the damage arrives fast. The leadership shortage takes years to become visible on an org chart, but the disorientation starts within months.
I advised a COO who inherited exactly this problem eighteen months after her company flattened its structure. Her best analysts kept repeating the same mistake, since nobody was catching it early anymore. She hadn’t lost talent. She had lost the layer that used to correct talent while the stakes were still small.
When we mapped out what had happened, the pattern was simple. Every analyst on her team was still technically capable. What they lacked was the weekly, low-stakes correction that used to come from a manager who noticed a flaw before it reached a client. Removing that manager didn’t lower the skill bar. It removed the mechanism that used to catch people underneath it.
The Skill Nobody Told You Middle Management Was Teaching
Here is the part most restructuring plans skip entirely. Middle management was never primarily a coordination function. It was a training ground for judgment under ambiguity, the specific skill of making a call with incomplete information and living with what comes next.
You cannot read your way into that skill. You cannot watch a webinar and absorb it. It gets built by making a decision that goes wrong in front of eight people, then making the next one better.
Philosopher Michael Polanyi called this kind of learning tacit knowledge, the understanding we carry but cannot fully explain, built through repetition rather than instruction. Leadership judgment behaves the same way. It transfers through lived experience, not through a slide deck.
A founder I worked with put it more bluntly than I would have. He told me his best vice president had been “seasoned by failure nobody wrote down,” meaning she made three bad calls early in her career that never appeared in a performance review but taught her more than any leadership program since.
When you remove the layer where those quiet failures used to happen, you do not save that development cost. You defer it, and it returns later as an underprepared executive making a far more expensive mistake in front of the board instead of a team of eight.
I’ve noticed the same pattern across nearly every executive I’ve worked with who rose through a traditional management track. None of them describe their best learning as a training program. Almost all of them describe a specific week when a decision went badly and someone senior let them own the fallout instead of stepping in to fix it. That ownership, not the coordination work, was the actual value the role provided.
What Do Companies Lose When They Cut Middle Managers?
Companies cutting middle managers lose the informal bench of future senior leaders who were being tested and corrected in real time. Only six percent of Gen Z employees now say they want a senior leadership role, according to Deloitte’s research, a direct consequence of watching that bench disappear.
They have watched companies eliminate the managers just above them, and concluded, correctly, that leadership is treated as disposable rather than developmental. That is not a generational attitude problem. It is a rational read of what they have observed.
Beyond that, companies lose the person who once caught a bad decision before it reached a customer or a board deck. Succession planning depends on someone having already made small, correctable mistakes in a role with real but limited consequences. Remove the role, and the mistakes simply move up the chain, arriving later and costing more.
There is also a quieter loss that rarely makes it into the analysis. Middle managers were often the people who noticed a high performer heading toward burnout, or a talented employee about to quit out of frustration, weeks before anyone else caught it. Losing that early-warning function does not show up on a balance sheet, at least not until the resignation letters do.
Can AI Replace Leadership Development and Coaching?
AI can absorb the coordination and reporting work middle managers used to carry, but it cannot replace coaching, trust-building, or judgment developed through lived accountability. Those capacities are built through repeated, consequential human interaction, not through a tool, however well designed the tool might be.
I’ve watched companies hand high-potential employees an AI leadership simulator as a substitute for the manager they cut. The tool can model a difficult conversation. It cannot make someone feel the actual discomfort of delivering bad news to a person whose mortgage depends on the outcome, and that discomfort is where judgment gets built.
This is also why the Gen Z statistic above matters so much. It is not that younger employees lack ambition. It is that they are watching the apprenticeship model disappear and drawing the obvious conclusion about what the organization values.
I raised this directly with a CMO I advised who had rolled out an AI coaching assistant across her marketing organization. She was proud of the adoption numbers. When I asked how many of her high-potential employees had made a genuinely consequential decision on their own in the past quarter, she went quiet. The tool had improved communication. It had not replaced the moment where someone stands alone with a hard call and must own the outcome.
The Real Cost Behind the Headcount Savings
One widely cited 2026 analysis estimated that the million companies save cutting middle managers today can cost roughly double within three years, once external hiring premiums and the fallout from underprepared promotions are added back in. The savings arrive immediately. The bill arrives later, usually after the leader who approved the cut has already moved on.
I am not arguing against flatter structures on principle. Some companies genuinely carried too many layers and removing them improved decision speed without damaging development at all. Pretending every flattening decision is reckless would be its own kind of false certainty, and false certainty is not something I want to sell any reader.
The difference is whether leaders replace what they removed. A structure can get flatter and still develop people, but only if someone deliberately rebuilds the mechanism for lived accountability instead of assuming it will happen on its own.
I think of it as a ledger, in the plain accounting sense. Every leadership cut has a credit column and a debit column. The credit shows up fast: lower payroll, tighter reporting lines, a cleaner org chart for the next board deck. The debit shows up slow, usually two or three years out, and it rarely gets traced back to the original decision. By the time the gap appears, the executive who approved the cut has often been promoted for the savings it produced.
How to Protect Your Leadership Pipeline Without Bloating the Org Chart
Rotational assignments with real budget authority do more for leadership development than a title change ever will. So do mentorship relationships where the mentor carries genuine stake in the mentee’s decisions, not a monthly coffee chat with no consequence attached.
Some analysts have started describing this emerging role as the “Meridian Manager,” a leader who translates organizational purpose into the behavior of increasingly autonomous systems while still holding the human judgment that governs both. Whatever the label, the function must exist somewhere in the structure, deliberately assigned rather than assumed.
A CEO I advised solved this differently than most. Instead of restoring the management layer she had cut, she gave six high-potential employees direct ownership of a project with real financial consequences, then met with each of them weekly to walk through what went wrong and why. Eighteen months later, three of them were running functions she would otherwise have hired externally to fill.
None of this required rebuilding the layer she had removed. It required her to stop assuming judgment would develop by osmosis, and to build a smaller, more deliberate version of the same apprenticeship instead. That took a real commitment of her own calendar, which is exactly why so few leaders do it. Cutting a layer is a one-time decision. Replacing what it taught is a standing obligation, and it competes with everything else on a CEO’s schedule.
Boards are starting to ask sharper questions about this too. A few now request a one-page summary of how leadership development continues after a flattening decision, alongside the projected savings, not instead of them. That is a healthy shift, and it puts pressure on executives to have a real answer ready before the meeting, not after the resignations start arriving.
That is the trade worth naming honestly. You can flatten the structure. You cannot flatten the requirement that someone, somewhere, learns to carry weight before the stakes get large.
If you are cutting layers this year, the real question is not whether you can afford fewer managers. It is who, specifically, is going to carry the weight those managers used to carry, and whether you have told them yet.
If this connects with something you are wrestling with, my novel The Ledger explores exactly this kind of reckoning: what a leader owes the people coming up behind them, and what gets inherited when nobody names it out loud. It’s available on Gumroad: https://5402694886686.gumroad.com/l/xhiax