March 12, 2026
By Vanguard Enterprise Intelligence Unit with the work of Amy Edmondson, Roger Martin, John Kotter, Herminia Ibarra, and Robert Cialdini.

Every organization has two power structures. The first is visible. It appears in reporting lines, titles, committees, budgets, approval rights, and formal governance. The second is informal. It appears in relationships, expertise, access to information, control over critical resources, reputation, trust, and the ability to persuade others to act. Most leaders manage the first structure carefully. The second often runs the company.

In stable environments, the gap between formal and informal power can remain tolerable. The hierarchy sets direction. Managers translate priorities. Teams execute. Informal networks help work move around the edges. But in turbulent environments, the gap becomes dangerous. Decisions must be made faster. Information must travel more quickly. Teams must coordinate across functions. Innovation must occur closer to the customer. Risk must be escalated before it becomes crisis. The old hierarchy alone cannot carry the load.

This is why organizational power is being redesigned. Companies are not simply flattening structures or adopting fashionable agile language. They are trying to solve a deeper problem: how to distribute authority widely enough to move quickly while preserving enough coherence to remain governable. Too much centralization creates delay. Too much decentralization creates fragmentation. Too much hierarchy suppresses initiative. Too much autonomy weakens accountability.

The task for leaders in 2026 is to build power systems that are both distributed and disciplined. The goal is not to abolish hierarchy. The goal is to make authority flow to the places where judgment, information, and accountability can be joined.

The Failure of the Old Power Map

Traditional organizational design assumed that power should follow the formal structure. Senior leaders made strategic choices. Middle managers translated those choices. Frontline teams executed. Support functions controlled risks, resources, and standards. The model worked reasonably well when markets were more predictable, work moved in linear sequences, and information traveled upward through stable channels.

That model now struggles. AI, geopolitical volatility, hybrid work, supply-chain instability, customer fragmentation, and regulatory complexity have made decision-making more distributed in practice, even when the formal structure remains centralized. The person closest to the relevant information may sit three layers below the executive team, in a different geography, or outside the company altogether. A formal leader may have authority without context. A frontline team may have context without authority.

This mismatch creates organizational drag. Decisions move upward for approval, then downward for execution, losing time and nuance in both directions. Cross-functional teams spend weeks negotiating ownership. Managers protect their areas because incentives reward local success. Employees wait for clarity that never arrives. Informal influencers make the work happen, but without formal recognition or accountability. The organization appears structured, but power is misaligned with reality.

The result is neither control nor agility. It is delay with the illusion of discipline.

Distributed Power Without Chaos

Distributed power is attractive because it promises speed, innovation, and responsiveness. If teams closer to customers, technology, operations, and local markets can make more decisions, the organization can adapt faster. But distributed power can easily become chaos if leaders fail to define boundaries.

The issue is not whether authority should be pushed downward. It should be, in many cases. The issue is which authority, under what conditions, with what constraints, and with what accountability. A product team may need authority to adjust features quickly, but not to ignore cybersecurity standards. A regional leader may need authority to adapt go-to-market strategy, but not to violate global compliance rules. An AI team may need autonomy to experiment, but not to deploy systems that affect customers without review. A procurement leader may need flexibility to secure supply, but not to bypass human rights due diligence.

Distributed power works only when the organization distinguishes between decisions that should be local, decisions that should be shared, and decisions that must remain enterprise-wide. Without that distinction, autonomy becomes inconsistency. With it, autonomy becomes speed inside a coherent system.

The best organizations do not simply decentralize. They specify the architecture of decentralization.

Decision Rights as the Infrastructure of Speed

Many organizations talk about moving faster while leaving decision rights vague. This is one of the most common causes of organizational friction. When people do not know who decides, who advises, who can veto, and who owns the result, speed collapses into negotiation.

Decision rights are the infrastructure of speed. They define where authority sits and how decisions move. They prevent every issue from becoming a meeting. They reduce the need for political interpretation. They give teams permission to act.

But decision rights must be designed carefully. Traditional tools such as RACI matrices can help clarify roles, but they often become bureaucratic if treated as checklists rather than management disciplines. The point is not to label everyone responsible, accountable, consulted, or informed. The point is to make decisions faster and better.

A useful decision-rights model begins with the type of decision. Strategic choices that affect enterprise direction require senior alignment. Operating decisions should often sit closer to execution. Risk-sensitive decisions require review by control functions. Customer-impacting decisions may require both speed and safeguards. AI-enabled decisions may require special clarity about whether a human, a system, or a hybrid process is actually shaping the outcome.

Leaders should ask five questions. Who has the best information? Who bears the consequences? Who controls the resources? Who must be consulted because they own risk? Who is accountable if the decision fails? Decision rights should not follow status alone. They should follow the relationship between information, consequence, and accountability.

Reporting Lines Are Not Enough

Reporting lines still matter. They define managerial responsibility, career development, resource allocation, and escalation. But they are insufficient for work that depends on networks. A product launch may require marketing, sales, technology, finance, legal, supply chain, and customer support. A transformation initiative may require HR, data, operations, cybersecurity, and external vendors. A sustainability commitment may require procurement, engineering, suppliers, investors, and regulators.

In this kind of work, the formal reporting line tells only part of the story. The more important question is how influence moves across the system. Who do people call when they need a decision? Who can unblock resources? Who has credibility with skeptical teams? Who understands both the technical and commercial dimensions? Who can translate between functions? Who is trusted by the frontline? Who can sense resistance before it becomes visible?

These informal networks often determine whether strategy succeeds. Yet many companies do not map them. They reorganize boxes on an org chart while leaving the actual flow of influence untouched. They change reporting lines but fail to change how decisions, information, trust, and expertise move.

Organizational redesign should therefore include network analysis. Leaders should identify informal connectors, bottlenecks, hidden experts, overloaded decision nodes, and trust gaps between functions. Sometimes the answer is not another structural layer. It is strengthening the connective tissue that already exists.

The Power of Informal Networks

Informal networks can be a source of agility. They help organizations bypass unnecessary bureaucracy, solve problems quickly, and spread knowledge across boundaries. A respected engineer can influence product decisions beyond his formal role. A regional manager can warn headquarters about local market risk. A finance partner can help teams understand tradeoffs before formal approval. A long-tenured operator can prevent a strategy from failing in execution.

But informal networks can also undermine accountability. If decisions are made through hidden channels, the organization may not know who owns the outcome. If influence depends too heavily on personal relationships, new talent may be excluded. If informal power concentrates in a few individuals, the company becomes vulnerable when they leave. If informal networks compensate for broken formal systems, leaders may mistake heroics for capability.

The goal is not to suppress informal networks. It is to make them visible and healthy. Leaders should recognize informal influencers, involve them in change efforts, and ensure that their influence supports enterprise priorities rather than local politics. They should also avoid building systems that rely on a few heroic connectors to compensate for unclear authority.

A healthy organization uses informal networks to accelerate trust, not to hide power.

The Case Pattern: The Overloaded Center

A common case pattern in organizational redesign is the overloaded center. A company wants agility but routes too many decisions through senior leadership, headquarters, or a central function. The center becomes the bottleneck. Every major customer exception, technology choice, talent approval, pricing deviation, risk question, and investment decision escalates upward. Senior leaders become overwhelmed. Teams become frustrated. Decisions slow down.

The center often defends this model by pointing to risk. If decisions are decentralized, leaders worry that standards will weaken. That concern is real. But centralization does not always create control. In many cases, it creates delayed control. Problems still happen, but later. Local teams still adapt, but informally. Workarounds multiply. The center becomes both too powerful and too uninformed.

The solution is not to eliminate the center. It is to redesign the center’s role. The center should set standards, define decision rights, manage enterprise risk, allocate capital, build shared platforms, and intervene on truly strategic issues. It should not approve every operational variation. A strong center creates clarity. A weak center hoards decisions.

The best organizations turn the center from a permission gate into a capability platform.

The Case Pattern: The Fragmented Edge

The opposite problem is the fragmented edge. In this pattern, business units, regions, or teams have too much autonomy without enough shared architecture. They choose different systems, define metrics differently, create inconsistent customer experiences, duplicate capabilities, negotiate separately with vendors, and interpret enterprise priorities through local incentives.

At first, the fragmented edge can look entrepreneurial. Teams move fast. Local leaders feel empowered. Innovation appears abundant. But over time, the enterprise loses coherence. Costs rise. Data becomes fragmented. Customers receive inconsistent treatment. Risk standards vary. Scaling becomes difficult because each part of the organization has built its own version of reality.

The solution is not to recentralize everything. It is to define what must be common. Common data standards, common risk controls, common customer principles, common technology architecture, common talent expectations, and common governance rules can coexist with local flexibility. The edge should have room to adapt, but not to reinvent the enterprise without discipline.

The art of organizational power is deciding what belongs at the center and what belongs at the edge.

Agile Structures and Accountable Governance

Agility is sometimes treated as the opposite of governance. That is a mistake. Poor governance slows organizations down. Good governance speeds them up because it clarifies boundaries before conflict arises.

An agile organization does not mean one where everyone can do anything. It means one where people can act quickly within understood rules. Product teams know what they can decide. Risk teams know when to intervene. Finance knows which thresholds require review. Executives know which metrics signal trouble. Employees know how to escalate concerns. Governance creates the guardrails that allow speed without recklessness.

This is especially important as AI enters more decisions. AI can accelerate analysis, recommendations, content generation, coding, customer service, and workflow execution. But it also shifts power. A decision may appear human-led when an AI system has framed the options. A manager may approve an action without understanding how the recommendation was generated. A team may automate a workflow before decision rights are clear.

Organizations must therefore design governance around human-AI decision systems. Who frames the problem? Who reviews the output? Who can override the system? Who is accountable for consequences? Which decisions can be automated? Which require human approval? AI makes organizational power less visible unless leaders deliberately make it visible again.

The Manager’s New Role

Managers often feel squeezed by organizational redesign. They are told to empower teams while remaining accountable for outcomes. They are told to move faster while protecting control. They are told to collaborate across functions while meeting local targets. They are told to use AI while preserving human judgment.

This tension is real. But it also defines the new managerial role. Managers are no longer merely supervisors of work. They are designers of decision environments. Their job is to ensure that people know what they can decide, what they must escalate, which stakeholders must be involved, and how success will be measured.

The best managers amplify their impact by building clarity around them. They clarify priorities. They clarify decision rights. They clarify tradeoffs. They clarify standards. They clarify how informal influence should be used. They do not try to control every action. They create conditions where the right actions are more likely.

This requires a shift in managerial identity. A manager’s power is no longer measured by how many approvals flow through him. It is measured by how effectively his team can act without unnecessary dependence on him.

A Framework for Redesigning Organizational Power

Leaders redesigning power systems should begin with strategic work, not structure. The first question is: what must the organization be able to do better? Move faster? Innovate more? Control risk? Serve customers more consistently? Scale AI? Enter new markets? Improve accountability? Structure should follow the capabilities the strategy requires.

The second question is: where should decisions sit? Leaders should map critical decisions and place them where information, expertise, and accountability are closest together. Decisions should move upward only when enterprise consequences, risk exposure, or resource commitments justify escalation.

The third question is: what must remain common? Common standards create coherence. These may include values, data definitions, risk thresholds, technology architecture, customer principles, AI governance, compliance rules, and financial controls. Without common standards, distributed power fragments.

The fourth question is: which networks matter? Leaders should identify the informal connectors and influence pathways that make work move. They should strengthen productive networks and reduce dependency on hidden bottlenecks.

The fifth question is: how will accountability be measured? Distributed authority without measurement becomes ambiguity. Teams need clear outcomes, leading indicators, review rhythms, and consequences. Accountability should be tied to both results and the quality of decisions used to produce them.

Strategies for Executives

Executives should begin by auditing decision friction. Where do decisions slow down? Where do issues repeatedly escalate? Where do teams wait for approval? Where do functions disagree over ownership? These friction points reveal where power is misdesigned.

They should then redesign governance around decision types. Not every decision needs the same process. Routine operating decisions should move quickly. Strategic decisions need alignment. High-risk decisions need review. Cross-functional decisions need clear ownership. AI-enabled decisions need transparency around human and machine authority.

Executives should also identify informal power brokers. These are the people others trust, consult, and follow. Some hold senior titles. Others do not. Bringing these actors into redesign efforts can accelerate adoption and reveal hidden realities that org charts miss.

Leaders should simplify reporting where possible, but they should not assume simplicity alone creates agility. A flatter structure can still be slow if decision rights are unclear. A matrix can work if governance is disciplined. The issue is not the number of boxes. It is whether power flows in ways that match the work.

Finally, executives should make accountability cultural. People should understand that autonomy is not permission to optimize locally at the expense of the enterprise. Distributed power works when leaders at every level feel responsible for both speed and stewardship.

The Real Redesign

The future organization will not be purely hierarchical or purely networked. It will be both. Hierarchy will remain necessary for accountability, capital allocation, leadership development, and strategic direction. Networks will remain necessary for speed, learning, innovation, and cross-boundary execution. The challenge is to design the relationship between them.

The companies that fail will swing between extremes. They will centralize after mistakes, then decentralize after delays, then recentralize after fragmentation. Their redesigns will be reactive, structural, and incomplete.

The companies that succeed will become more deliberate. They will distribute authority where information is strongest. They will centralize standards where coherence matters. They will make informal networks visible. They will build governance that enables speed. They will teach managers to design decision environments rather than merely supervise tasks.

Organizational power is not a problem to be eliminated. It is a system to be designed.

In turbulent times, the advantage belongs to companies that can move quickly without losing control, innovate without losing discipline, and distribute authority without losing accountability.

That is the new architecture of power.