The Great Flattening: Why Companies Are Cutting Middle Managers and What HR Leaders Are Missing
.png)
- Companies are flattening middle management because AI is taking over coordination work, but that creates new risks if coaching, context, and judgment are not replaced.
- The role of the manager is shifting from coordinator to orchestrator, with more focus on coaching, translating strategy, and building team conditions for human-AI collaboration.
- HR and L&D should redesign manager roles before cutting layers, then track hidden costs like engagement, onboarding speed, and knowledge transfer.
Manager headcount at public companies fell 6.1% between 2022 and 2025. Meta, Amazon, Google, Intel, and Estee Lauder are all flattening their hierarchies. The pitch is faster decisions and lower cost. The hidden cost is steep: 37% of employees say they feel directionless, and nearly half of senior executives doubt they can manage what's left. Flattening doesn't eliminate the work middle managers did. It redistributes it, and usually badly.
The strategic question is not whether to flatten. It's what the manager role becomes when AI handles the coordination work that justified the layer in the first place.
What the data actually shows
A structural shift is happening across corporate hierarchies, and the numbers are sharper than most HR leaders realise.
- Between May 2022 and May 2025, manager headcount at public companies fell 6.1%, while executive roles fell 4.6% (CNBC, 2025)
- Estee Lauder announced cuts of around 20% of its management workforce (CNBC, 2025)
- Meta, Intel, and Google have all publicly reduced their management layers, with Google cutting roughly a third of its managers (CNBC, 2025)
- Amazon CEO Andy Jassy has been explicit, saying he plans to cut middle managers who "want to put their fingerprint on everything" (Yahoo Finance, 2025)
The C-suite message is consistent: flatten, empower individual contributors, move faster. AI is the enabler. Coordination and reporting work that once required a layer of managers can now be handled by AI tools.
The pitch is real. The question is whether the savings are.
Why companies are flattening and what the headlines miss
The logic for flattening is straightforward. AI absorbs routine coordination, status reporting, and synthesis work. The administrative justification for multiple management layers shrinks. Fewer layers should mean faster decisions, less bureaucracy, and budget freed up for technical talent and AI investment.
What the headlines miss is what happens to the work that didn't go away.
When middle managers disappear, two things happen at the same time. Nearly half of senior executives now doubt their ability to manage everything that's landed on them, and 37% of employees say the lack of managers has left them feeling directionless (Korn Ferry, 2025).
The pattern is consistent across organizations that flatten without redesigning:
- Senior leaders drown in operational detail instead of focusing on strategy
- Junior employees lose the mentorship and context translation they need to grow
- AI fills the task layer, but not the human layer: connection, coaching, judgment, escalation
Removing the middle layer doesn't eliminate the work. It redistributes it, usually to people who weren't designed to do it.
What the new manager role actually looks like
The flattening trend is real, but the simplistic version of "AI replaces middle managers" misreads what's happening. The role isn't disappearing. It's being rewritten.
The shift is from coordinator to orchestrator. The manager who survives the flattening is the one who can compose human and AI capability into outcomes the business needs.
That sentence is the strategic point. Flattening removes the work AI can do. What's left is the work that creates value — coaching, collaboration, strategic thinking, organizational capability building. Managers who lean into that work become more valuable. Managers who don't, become exactly the headcount line item the C-suite is targeting.
The skills that define the new middle manager
The list of capabilities is shorter and harder than the old job description.
- Orchestrating human-AI collaboration: designing how teams and AI agents work together, deciding what humans own and what AI handles, redesigning workflows accordingly
- Coaching at depth, not at scale: moving from performance reviews and status updates to genuine development conversations
- Translating strategy into context: taking ambiguous executive direction and turning it into specific, motivating work for individual contributors
- Building team conditions: psychological safety, role clarity, and dependability — the structural conditions that make hybrid human-AI teams work
- Knowing what to escalate: judgment about which decisions go up, which stay with the team, and which the AI is authorised to make
These are not the skills the average middle manager was hired for. They are not the skills most management training programs teach. The gap between what's required and what's developed is the actual crisis HR should be solving.
What HR leaders should do before flattening
Flattening isn't inherently right or wrong. The question is whether the organization has built the infrastructure to replace what middle managers provided.
1. Diagnose before you cut
Map what each management layer actually contributes:
- Coordination and status work (likely automatable)
- Coaching and development (very hard to automate)
- Context translation between strategy and execution (hard to automate well)
- Knowledge transfer and onboarding (often invisible until it's gone)
- Judgment on escalations and trade-offs (still requires human judgment for most decisions)
If a layer's primary value is coordination, AI can probably absorb it. If its value is coaching, judgment, or context, removing it without replacement creates exactly the directionless feeling the data shows.
2. Redesign the role, not just the headcount
The remaining managers need a different job, not the same job with more direct reports.
Update:
- The job description (what the role is actually for)
- The skills profile required at hire and promotion
- The development support: most current managers were not trained to orchestrate AI-augmented teams
- The performance metrics: measuring orchestration outcomes, team conditions, and capability development, not just delivery
3. Watch for the hidden costs
Track more than the cost saving. The numbers that matter alongside it:
- Employee engagement and retention in flattened teams vs. comparable structures
- Time senior leaders spend on operational versus strategic work
- Internal mobility and promotion rates from individual contributor to manager
- Speed of onboarding and ramp-up for new hires
- Knowledge transfer indicators (who's training the next generation?)
Flattening that lifts these metrics is working. Flattening that drops them is producing a short-term cost saving in exchange for a long-term capability problem.
The strategic point
The Great Flattening is not the end of middle management. It's the end of middle management as it was designed in the 1990s: coordination layers built for an industrial-era information flow.
The version of middle management worth preserving is the human-development, capability-building, judgment-applying version. That version requires investment in different skills, different training, and different role design. It cannot be cut to a third without consequence.
The companies that flatten well will not be the ones with the most layoffs. They will be the ones that figured out what middle managers were actually for, automated the parts that should be automated, and rebuilt the rest as a sharper, more strategic role.
That is a workforce design problem. It is exactly the kind of problem strategic L&D should be at the center of.
Frequently Asked Questions
What is The Great Flattening?
The Great Flattening is the term used to describe the recent trend of organizations significantly reducing their middle management layers. Between May 2022 and May 2025, manager headcount at public companies fell 6.1%. Companies including Meta, Amazon, Google, Intel, and Estee Lauder have publicly cut large portions of their management workforce, citing AI-driven efficiency and the goal of empowering individual contributors.
Why are companies cutting middle managers?
Three reasons drive most decisions: AI tools can absorb routine coordination, status, and reporting work that justified manager layers; flatter hierarchies promise faster decision-making and reduced bureaucracy; and the cost savings can be redirected toward technical talent and AI investment. The C-suite logic is real, but the savings often hide capability costs that surface later.
What are the hidden costs of flattening hierarchies?
When middle layers are removed without redesigning what's above and below, the work doesn't disappear — it gets redistributed. Nearly half of senior executives report doubting their ability to manage everything left to them, and 37% of employees report feeling directionless. Other hidden costs include lost mentorship for junior staff, slower onboarding, and erosion of organizational knowledge.
How is the role of middle managers changing in the AI era?
The role is shifting from coordination to orchestration. Managers who used to coordinate work, collect status updates, and translate between layers now need to design how human teams and AI capabilities work together. The skill profile is moving toward coaching, strategic translation, building team conditions, and judgment about human-AI workflow design.
What skills do middle managers need in 2026?
The most important skills for the new middle manager role include orchestrating human-AI collaboration, coaching for development rather than performance management, translating strategy into team-level context, building psychological safety and role clarity, and judgment about decision escalation. These are different skills from those most middle managers were originally hired for.
Should companies flatten their hierarchy?
It depends on whether the organization is willing to redesign the work, not just remove the headcount. Flattening that combines AI automation with redesigned manager roles, redistributed responsibilities, and investment in new manager skills can work. Flattening that just removes the layer tends to produce a short-term cost saving and a longer-term capability problem.
What's the difference between flat and flattened organizations?
Flat organizations are designed from the start with minimal management layers, clear roles, and autonomous teams. Flattened organizations are traditional hierarchies that have had layers removed, often without redesigning the underlying work. Flat structures can work well. Flattened structures often inherit the worst of both worlds — fewer managers but the same expectation of how work gets coordinated.

We offer a scalable employee training solution. It lets you continuously upskill your people.
Book a call
"In 10 to 15 years, HR will look completely different, with a new skill set and a much broader scope. It will become a more business-led function, driving specific value areas toward execution. While some existing responsibilities will remain, they will play out in a very different way. HR will focus more on sustainability, the future of work, leadership capacity building, and talent and experience—rather than just transactional tasks. These tasks will still be important, but they will be handled differently. To meet future mandates and expectations, HR will need a fundamentally different skill set."

"The traditional command, control, and delegate cycle didn't work anymore because AI could handle those tasks. This leaves managers with a choice: use that freed-up time for strategy, collaboration, and helping teams develop, or become redundant."

Related articles
Ready to drive impact together?
Close skill gaps, accelerate growth, and future-proof your workforce.



.jpg)
.png)
.png)
.png)