← Library · Great Flattening
Great Flattening · 6 min read

Middle Management Is Shrinking. What Survives Must Change.

AI and cost pressure are compressing the middle of firms, but cuts create value only if companies redesign how coordination, coaching, and decisions happen closer to the work.

RTO mandates are back, AI pilots are multiplying, and budgets are tight. In many firms, the squeeze lands most painfully in the middle. U.S. employers were advertising 42 percent fewer midlevel manager roles at the end of 2024 than in spring 2022 (Deloitte, 2025). At the same time, manager engagement has slid sharply, falling from 27 to 22 percent in a single year, part of a wider global engagement slump to 20 percent that Gallup estimates drains about 9 percent of world GDP in lost productivity (Gallup, 2025). Boards want flatter structures and faster decisions. Employees want less bureaucracy and more agency. The risk is obvious. Take out layers without redesigning the work, and you get chaos rather than speed.

Why the squeeze is structural

Managers are spending their time in the wrong places. According to Deloitte’s 2025 Global Human Capital Trends, managers report spending nearly 40 percent of their time on firefighting and administrative work and only 13 percent on developing their people. More than a third say they feel unprepared for the people side of the job (Deloitte, 2025). Those numbers square with how leaders see their organizations. McKinsey’s State of Organizations 2023 found that two thirds of business leaders view their organizations as overly complex and inefficient, and only half believe their firms can anticipate and respond to shocks effectively (McKinsey, 2023).

Technology will accelerate the trend. Gartner forecasts that by the end of 2026, one in five organizations it tracks will use AI to flatten their structures (Deloitte, 2025). A related Gartner call, reported by Fortune, projects that in those companies more than half of current middle management roles will be eliminated (Fortune, 2025). These are forecasts, not outcomes. They do, however, capture a direction of travel. If AI can route information, summarize decisions, and standardize workflows, it erodes the traditional value of the person in the middle who knew where things were and who to ask.

The danger is mistaking subtraction for strategy. Simply deleting roles will not lift engagement or performance. Gallup’s comparison between typical and best-practice organizations is telling. In top-quartile organizations, 79 percent of managers are engaged, versus 22 percent on average, a gap that points to design choices rather than fate (Gallup, 2025). The work of managing can be made meaningful. It will not happen by memo.

Flattening needs a mechanism, not a memo

Debates about self-management often polarize into romance or ridicule. Ethan Bernstein and colleagues examined several high-profile experiments and urged a balanced reading of what self-organization can and cannot do. As of 2016, Zappos was the largest company to attempt a full adoption of holacracy, a detail that grounds the conversation in practice rather than theory (Bernstein, Bunch, Canner and Lee, 2016). Their central point holds up today. Structures that distribute authority need mechanisms that make accountability, coordination, and learning visible.

McKinsey highlights firms that try to do this at scale, naming Haier as a beacon for a self-evolving model with micro-enterprises that push decisions closer to customers (McKinsey, 2023). Post-pandemic, about 90 percent of organizations have adopted some form of hybrid work. That shift increases the premium on clear interfaces, documented processes, and outcome-based management. Without those, hierarchy grows back to plug the gaps.

Where no middle managers work today

The cleanest contemporary case is Buurtzorg, the Dutch home-care organization that scaled to more than 900 self-managing teams with roughly 14,700 employees as of January 2021 and no middle managers or supervisors. The company operates with only two directors, about 50 back-office staff, and a cadre of 21 internal coaches who support roughly 950 teams, about one coach for more than 45 teams, and who hold no formal authority and are not responsible for team productivity (Martela and Nandram, 2025). Buurtzorg reported an overhead of about 8 percent versus a sector average of 25 percent in a KPMG assessment around 2015. That comparison is dated, and Buurtzorg’s cost per hour has been higher than peers, €54.47 versus €48.74, which underlines the need to read the model as a design choice rather than a universal cost cure (Martela and Nandram, 2025).

The case also carries boundary conditions. Buurtzorg’s approach works best where interdependence between teams is low, frontline professionals are both motivated and highly competent, and the service lends itself to local autonomy. In other words, removing middle managers is safer when the unit of work can be owned end to end by a small team, the outcomes are legible, and escalation is the exception rather than the rule (Martela and Nandram, 2025).

A practical redesign for 2025

Middle management is not ending. The role is splitting. Some tasks migrate to platforms and AI. Others become more human and harder to fake. Four moves help firms make that split deliberately rather than by attrition:

  • Reclaim manager time for coaching. Audit what drives the 40 percent firefighting and administrative load, then automate or centralize it. Track the share of time spent on developing people and make it a management metric. The baseline is low at 13 percent, which signals headroom (Deloitte, 2025).
  • Widen spans with scaffolding, not slogans. If a manager’s span grows, teams need clear operating cadences, decision rights, and transparent metrics. Where interdependence is low and expertise is high, pilot self-managing teams with coach support and no formal supervisory authority, as Buurtzorg does. State upfront where the model will not apply due to tight cross-team coupling (Martela and Nandram, 2025).
  • Treat AI as a coordinator. Use AI to route requests, summarize context, and propose next steps so decisions can move to the edge. Gartner’s forecast suggests many firms will push this way. The risk is that eliminating roles outpaces the maturity of the operating system. Phase changes with explicit service levels and escalation paths (Deloitte, 2025; Fortune, 2025).
  • Manage for engagement as a design outcome. In the best organizations, nearly four out of five managers are engaged. That gap with the 22 percent global average suggests that structure, clarity, and purpose can be engineered. Tie manager incentives to team engagement and performance, and use engagement data as an early warning against complexity creep (Gallup, 2025).

The transition will not look the same everywhere. In complex, tightly coupled businesses, cut less and equip managers to be integrators who translate across functions and time zones. In modular, service-centric contexts, push more authority into stable, small teams and replace supervisory checks with peer review and customer outcomes. Either way, the message is the same. The end of the middle manager is not about erasing a pay grade. It is about stripping away coordination work that software and clear rules can do and elevating what remains into coaching, judgment, and stewardship. Firms that make that trade consciously will flatten without falling flat.

Sources

  • Deloitte (2025). 2025 Global Human Capital Trends. Finds managers spend ~40% on firefighting/admin and only 13% on people development, notes a 42% drop in U.S. midlevel manager job postings since 2022, and cites a Gartner forecast that 20% of organizations will use AI to flatten structures by 2026. https://www.deloitte.com/us/en/insights/topics/talent/human-capital-trends/2025.html
  • Gallup (2025). State of the Global Workplace 2025. Reports global engagement at 20% with an estimated cost near 9% of global GDP, and a sharp drop in manager engagement from 27% to 22% between 2024 and 2025, while best-practice organizations have 79% of managers engaged. https://www.gallup.com/workplace/659610/state-global-workplace.aspx
  • Fortune (2025). Surviving the great flattening: the coming extinction of the middle manager. Cites a Gartner forecast that by end-2026, 20% of organizations will use AI to flatten and in those firms more than half of middle management roles will be eliminated, stated as a projection not a present fact. https://fortune.com/2025/09/19/surviving-great-flattening-coming-extinction-of-middle-manager-layoffs/
  • Martela and Nandram (2025). Buurtzorg: scaling up an organization with hundreds of self-managing teams but no middle managers. Documents Buurtzorg’s scale without middle managers, the coach model without formal authority, dated overhead comparisons around 2015, higher hourly costs, and boundary conditions for where the model fits. https://link.springer.com/article/10.1007/s41469-024-00184-y
  • Bernstein, Bunch, Canner and Lee (2016). Beyond the Holacracy Hype. Argues for a balanced view of self-management experiments and notes that Zappos was, at the time, the largest firm attempting full holacracy. https://hbr.org/2016/07/beyond-the-holacracy-hype
  • McKinsey (2023). The State of Organizations 2023. Reports that two thirds of leaders see their organizations as overly complex and inefficient, only half feel prepared for shocks, and notes the broad adoption of hybrid work while highlighting Haier as a beacon of a self-evolving organization. https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/the-state-of-organizations-2023

Produced by the Semco³ content engine · references cited inline · published first on this platform.

More on this theme