Cutting Management Layers Isn't the Same as Cutting the Work

Korn Ferry asked 15,000 professionals across roughly fifteen countries whether their company had trimmed management layers in the past year. Worldwide, 41% said yes. In the US specifically, it was 44%, and four in ten US workers said they now feel a lack of direction at work. Those two numbers came out of the same survey, and I don't think that's a coincidence.

Here's the pattern I keep seeing when a company announces it's "flattening": the org chart gets a new shape, a press release explains the thinking, and somewhere in the second paragraph is a phrase like "simplifying the structure." What's almost never in that press release is an account of what happens to the coordination work the removed layer used to do. Not the title. The actual work – the 1:1s, the cross-team negotiation, the "wait, whose call is this" conversations. Somebody was doing that. Deleting the title doesn't delete the task.

The vocabulary is doing more work than the org chart

Look at how the recent round of cuts got described. TechCrunch's running tally of 2026 layoffs has GitLab cutting about 350 people – 14% of staff, exiting 22 countries – framed internally as a "generational rebuild" that included flattening management layers. Intuit cut roughly 3,000 jobs, 17% of its workforce, described as "reducing complexity and simplifying the structure." PayPal's CEO framed a plan to cut 4,500-plus jobs over two to three years, about a fifth of the company, explicitly as "removing organizational layers," tied to a function-by-function AI redesign.

Every one of those is a real, verifiable move, and I'm not arguing the cuts were unjustified. What I'm pointing at is narrower: none of the public framing for any of them describes a parallel redesign of how the coordination work gets done once the layer is gone. "Simplifying the structure" describes the chart. It doesn't describe the mechanism that used to keep six teams pointed at the same priority.

Not every company reaches for that language, either, which is its own tell. IBM's cuts were described as "routine rebalancing affecting a low single-digit percentage of its global workforce" – notably not "simplification," a much more modest framing for a much more modest move. And Cloudflare's CEO talked about cutting roughly a fifth of the company, weighted toward middle management, finance, legal, and audit – real, and management-heavy, but a broad headcount reduction, not a stated delayering initiative. Worth knowing the difference: some of what gets read as "the flattening trend" is actually several different things wearing the same headline.

Where the work actually goes

It doesn't vanish. It moves down, onto whoever's left, without a system built to absorb it. Perceptyx's research on post-layoff organizations found employee engagement dropping from 51% to 44% after cuts. Onwards HR found confidence in leadership falling 10.5 percentage points after a reduction in force. Neither of those numbers is about the layoff itself. They're about what happens afterward, when the same amount of coordination has to route through fewer people who were never given new authority, new tools, or new time to do it.

That's the other half of Korn Ferry's finding: 40% of US workers now say they feel a lack of direction at work. Not "we have fewer managers." A lack of direction – the actual output managers were supposed to produce. If you flatten the chart and direction-giving doesn't degrade, you redesigned the mechanism. If you flatten the chart and direction-giving does degrade, you didn't redesign anything. You just relocated the burden and called it progress.

This isn't a morale problem. It's a delivery problem.

Most of what gets written about post-layoff organizations treats the fallout as a culture issue – engagement scores, sentiment surveys, exit interviews. That's real, but it's downstream of something more mechanical, and the mechanical version is the one I'd actually go looking for if I ran engineering or product at a company that just flattened.

It shows up as ticket cycle time creeping up for reasons nobody can name. It shows up as the same decision getting re-litigated in three different Slack threads because nobody's sure who owns it anymore. It shows up as rework – not because the work was done badly, but because it was done without the cross-team context a manager used to broker, so two teams built incompatible things and found out at integration time. None of that appears on a headcount slide. All of it is measurable, if anyone's looking.

I've made a version of this argument before about process change more broadly: three decades of adopting new methodologies hasn't meaningfully moved the industry's software delivery success rate, because swapping the label on how work gets organized isn't the same as changing the underlying mechanism that determines whether it ships well. Flattening the management layer is the same move at the org-structure level instead of the process level. And the "who absorbs the work" question isn't new to this moment, either – it's the same downward-pressure pattern I've pointed to in what happens to junior work when AI takes the easy 80%: the work doesn't disappear just because the person who used to do it is gone. It lands on whoever's left, usually without anyone deciding that's what should happen.

The tell that separates the two

I'd trust a flattening initiative that can answer three questions. Who now makes the calls a manager used to make, and do they know they're making them? Whose calendar absorbed the 1:1s, the mentoring, the cross-team friction – and is that time accounted for anywhere, or is it just happening on top of their existing job? And is anyone tracking decision latency – how long it takes to get an answer to something that used to go to a manager – before and after the change?

Most organizations cutting management layers right now can't answer any of those three. That's not a moral failing. It's a measurement gap, and it's the same gap I've written about in why a prioritization framework can look rigorous and still not change how decisions get made – a process artifact that looks like a fix but isn't touching the actual mechanism underneath it. Cutting a management layer is a structural change. Treating it as one requires redesigning the structure, not just shortening it.

What this actually is

I don't think most of this is cynical. I think a lot of leadership teams genuinely believe that removing a layer of managers removes a layer of overhead, full stop – that the coordination work was mostly the manager justifying their own existence, and once they're gone, the work goes with them. Sometimes that's true. Often it isn't, and the org finds out three months later when decisions that used to take a day are taking two weeks, and nobody can point to when that started because nobody was measuring it in the first place.

Cutting a management layer can be the right call. Cutting a management layer and assuming the coordination work disappeared with the title is a bet, not a strategy – and right now, most companies making that bet aren't even tracking whether they won it.