The "Great Flattening" Is Happening — But Should It?
There is a quiet revolution reshaping the modern workplace, and it has a name: the Great Flattening. Across industries, companies are stripping away layers of management in the name of efficiency, agility, and cost reduction. The logic sounds reasonable on the surface — AI can automate scheduling, performance tracking, goal-setting, and reporting, so why do we need a human layer in the middle? But a closer look at the data tells a far more complicated, and cautionary, story.
According to a Korn Ferry survey of more than 15,000 professionals worldwide, 41% of employees reported that their companies trimmed management layers in the past year alone. Middle managers accounted for more than 31% of all white-collar layoffs in 2023, and Gartner predicts that through 2026, 20% of organizations will use AI as justification to eliminate more than half of their middle management positions. The trend is not slowing down — it is accelerating.
Yet the question that rarely gets asked loudly enough is this: what exactly are we losing when we eliminate the middle manager role?
What Middle Managers Actually Do (That AI Cannot)
Part of the problem is that middle managers are chronically undervalued in public discourse. They are often portrayed as bureaucratic bottlenecks — people who sit in too many meetings, pass information up and down the chain, and add friction to otherwise efficient systems. CEOs dreaming of self-managing teams tend to focus on this caricature rather than on the reality.
The reality is that middle managers perform a set of deeply human functions that no dashboard or AI model has yet replicated at scale. These include:
- Building psychological safety: Great middle managers create environments where team members feel safe to take risks, admit mistakes, and offer honest feedback. This kind of trust is built through consistent human interaction, not algorithmic nudges.
- Translating strategy into meaning: Senior leadership sets direction. Middle managers translate that direction into something personally relevant for each team member, adjusting communication style, context, and emphasis depending on the individual.
- Navigating ambiguity: When priorities shift, when projects stall, or when conflict arises between colleagues, middle managers make judgment calls that require empathy, organizational knowledge, and relational history — none of which AI possesses.
- Developing talent: Career growth rarely happens through performance management software alone. It happens through a manager who notices potential, advocates for opportunities, and invests time in coaching and feedback conversations.
- Absorbing organizational stress: Middle managers act as buffers. They protect their teams from the chaos and noise of senior leadership while simultaneously translating ground-level concerns upward. This buffering function is invisible until it disappears.
The Self-Managing Team Fantasy
The appeal of self-managing teams is real. Companies like Spotify, Valve, and Buurtzorg have generated significant buzz with their experiments in radical autonomy and flat structures. But even these organizations have discovered that coordination, accountability, and interpersonal conflict do not simply dissolve when formal management titles are removed. Someone still ends up doing the work of a manager — they just do it without recognition, without authority, and often without compensation.
Research consistently shows that teams without clear leadership structures tend to suffer from role ambiguity, unresolved conflict, and inequitable workload distribution. High-performing self-managing teams are the exception, not the rule, and they typically require exceptional levels of team member maturity, psychological safety, and organizational support — conditions that take years to build and that rarely exist at scale.
Removing management layers without replacing the functions those managers performed is not organizational evolution. It is organizational subtraction dressed up as innovation.
AI Augments Management — It Does Not Replace It
The argument that AI will make middle managers obsolete conflates two very different things: administrative tasks and human leadership. Yes, AI can take meeting notes, generate performance summaries, schedule one-on-ones, and flag anomalies in productivity data. These are real efficiency gains, and organizations should absolutely embrace them.
But automating the administrative layer of management is not the same as automating the leadership layer. The moments that define whether a manager is truly effective — the difficult conversation about performance, the mentor relationship that changes the direction of someone's career, the team crisis navigated with calm and clarity — these are irreducibly human moments.
The most productive framing is not "AI versus the middle manager" but rather "AI empowering the middle manager." When managers are freed from administrative burdens, they can invest more time in the high-value, relationship-intensive work that actually drives team performance. Organizations that understand this distinction will outperform those that simply count the cost savings from cutting headcount.
The Hidden Cost of Flattening
When middle management layers are cut without a thoughtful transition plan, several predictable problems emerge. First, individual contributors lose their primary advocate within the organization. Second, senior leaders become overwhelmed with operational detail that previously did not reach their level, compromising their ability to think strategically. Third, employee engagement and retention tend to decline, particularly among high performers who valued development and clear career pathways.
These are not hypothetical risks. Research from Gallup has repeatedly found that the quality of an employee's direct manager is the single strongest predictor of employee engagement — stronger than pay, benefits, or company culture in the abstract. Cutting that relationship in the name of efficiency is a trade-off that rarely looks favorable once the full costs are counted.
What Organizations Should Do Instead
Rather than eliminating middle management wholesale, smart organizations are rethinking what they ask of middle managers. This means reducing administrative overhead through AI tools, increasing manager-to-employee ratios thoughtfully, investing in manager training and development, and redefining the manager role around coaching and culture rather than compliance and control.
The goal should be better middle management, not less of it. The data does not support the elimination of this role — it supports its evolution. Organizations that ignore that distinction may find, too late, that they have flattened their way into a performance crisis.