When every decision still needs permission, you’ve stopped scaling

Companies love the language of empowerment. They talk about ownership, accountability and autonomy as though these are inherent qualities of high-performing teams. Yet the moment someone makes an important decision, five people suddenly need to approve it.

That is one of the clearest signals that an organisation has stopped scaling.

You can keep adding teams, managers, functions and processes, but if meaningful decisions still need to travel upwards, you are not creating leverage. You are creating a larger organisation competing for access to the same small group of people.

The company may be growing. That does not mean it is scaling.

Headcount is not scale

Most organisations still confuse growth with scale. More engineers, more designers, more product managers and more managers create the appearance of greater organisational capacity. The company gets bigger, more specialised and more complex, so it feels as though it must also be becoming more capable.

But scale is not simply the ability to add people. Scale is the ability to increase what the organisation can achieve without increasing coordination, approval and management overhead at the same rate.

One of the simplest ways to test this is to look at decisions. As the organisation grows, can more good decisions happen without more of them travelling upwards? If the answer is no, the organisation is not scaling its capability. It is building a larger queue.

This distinction matters because a company can increase headcount significantly while becoming slower at exactly the same time. More people create more work, more dependencies and more decisions. If decision-making remains concentrated in the same small group, those additional people do not create proportional leverage. They create additional demand on an already constrained system.

The operating model that got you here will eventually slow you down

Centralised decision-making is not inherently bad. In the early stages of a company, it often makes complete sense.

Founders usually hold most of the context. The organisation is small enough for information to travel quickly, and many decisions have outsized consequences. When everyone can fit in the same room, direct involvement from a founder or senior leader can make decisions faster and better. The problem is that organisations often keep this operating model long after the conditions that made it effective have disappeared.

As companies grow, the number of decisions increases dramatically. There are more products, customers, countries, functions, edge cases, regulations, dependencies and competing priorities. Teams are working on problems that senior leaders cannot possibly understand in the same depth as the people closest to them. Yet the decision model often remains largely unchanged.

Teams prepare a recommendation and wait. A meeting is scheduled. Someone asks for more context. The work is revised. Another stakeholder appears. A pre-meeting happens before the actual meeting so everyone can align on what they are going to say. The decision is eventually made by someone senior enough to provide permission. Then the organisation calls this alignment. A lot of the time, it is simply latency.

The problem is not that people are collaborating. Collaboration is necessary. The problem begins when collaboration becomes a substitute for clear authority and every consequential decision starts requiring organisational consensus.

Permission is an invisible tax

Most companies have very little understanding of how much time they lose waiting for decisions.

Permission rarely appears in project plans. Nobody creates a roadmap item called “three days waiting for executive approval”. Nobody includes “seven stakeholder reviews” in the estimate for shipping a feature. Yet these delays exist across almost every part of the organisation.

A hiring manager loses a strong candidate because compensation requires another approval. A designer stops progressing a direction because a senior stakeholder has not seen it yet. A product manager delays a decision because experience tells them that it will probably be reopened later. An engineering team creates a temporary workaround because nobody wants to take responsibility for the underlying decision.

None of these moments looks particularly serious in isolation. Collectively, they create enormous organisational drag.

This is how companies can add significant headcount without feeling noticeably faster. They have increased execution capacity without increasing decision capacity.

The result is an organisation in which work spends surprisingly little time being worked on and a remarkable amount of time waiting.

At some point, the bottleneck is the operating model

Leaders rarely set out to become bottlenecks. Most centralised decision-making starts with reasonable intentions.

Leaders want consistency. They care about quality. They want to ensure that the strategy is being followed and that teams are not making avoidable mistakes. Staying close to decisions can initially improve outcomes, particularly when a company is moving quickly and building new organisational capability. The problem begins when involvement turns into dependency.

At that point, teams start behaving differently. Instead of asking what the right decision is, they begin asking what decision is most likely to get approved. That shift is subtle, but it changes the organisation. The customer is no longer the only audience for the work. The approval system becomes an audience too.

Teams start learning how particular executives think. Managers become good at packaging recommendations. Senior practitioners become translators between the work and the people who hold authority. Meetings increasingly focus on navigating stakeholders rather than improving the quality of the decision itself.

Eventually, leaders become frustrated that people are not demonstrating enough ownership. But ownership is difficult to develop in a system that repeatedly removes it.

If you hire senior people and still require them to get permission for the decisions they were hired to make, you have not really hired leaders. You have hired expensive operators.

Organisations can train people not to make decisions

Decision-making is a capability, and like any capability, it improves through practice.

People build judgement by making decisions, understanding the trade-offs, seeing the consequences and adjusting their thinking over time. Strong organisations create enough space for people to exercise judgement while providing the context and constraints that make good decisions more likely. When decisions are repeatedly escalated, the opposite happens.

People stop developing strong points of view because the cost of doing so no longer feels worthwhile. They bring several options rather than making a recommendation. They ask what leadership prefers. They avoid reasonable risk because escalation is safer. They wait for certainty because acting without permission can create political consequences.

Eventually, leadership looks at this behaviour and concludes that people are not senior enough to be trusted with important decisions. The system has created the behaviour it now uses to justify the system.

This is one of the more destructive loops in scaling organisations because it steadily removes capability from the people closest to the work while reinforcing the belief that decision-making needs to remain centralised.

Accountability without authority is theatre

Organisations often try to solve these problems by increasing accountability.

They create clearer goals, stronger OKRs, tighter performance expectations and more explicit ownership. Leaders tell teams to act like owners and take greater responsibility for outcomes. But accountability means very little if the authority required to deliver those outcomes sits somewhere else.

If a team owns conversion but cannot change the onboarding experience without approval, their ownership is conditional. If a design leader is accountable for product quality but cannot set or enforce the quality bar, their accountability is incomplete. If a product leader owns a market but pricing, packaging, positioning and roadmap decisions are continually escalated, they are coordinating the business rather than truly owning it.

This happens because organisations often delegate responsibility much faster than they delegate authority. The language changes before the power structure does.

People become accountable for outcomes, but the decisions that shape those outcomes remain distributed across executives, committees and functions. When performance drops, the team is asked to explain the result even though the most consequential trade-offs were made somewhere else.

That is not meaningful accountability. It is responsibility without control.

Autonomy does not mean chaos

One of the most common arguments against distributing decision-making is that greater autonomy will lead to inconsistency.

If teams have more authority, the thinking goes, everyone will simply do whatever they want. Products will fragment, standards will fall and leaders will lose control. But that is not autonomy. That is an absence of leadership. Strong organisations do not replace control with chaos. They replace constant approval with better constraints.

Strategy tells teams what matters. Principles help people navigate trade-offs. Design systems establish experience standards. Technical architecture creates engineering boundaries. Budgets define financial constraints. Metrics make outcomes visible. Decision frameworks clarify who has the authority to make particular types of decisions.

These mechanisms allow organisations to distribute decisions without abandoning coherence.

This is a different form of leadership from reviewing everything personally. Instead of being the person who makes or approves every decision, the leader designs the environment in which good decisions can happen without them. That is much harder than approval. It is also much more scalable.

Push decisions towards information

Organisational hierarchy creates a strange tension. As authority increases, proximity to the work often decreases.

A researcher who has spoken to twenty customers knows things an executive does not. The engineer maintaining a platform understands constraints that will never appear in a strategy deck. The designer working through a workflow sees inconsistencies that senior leaders may never encounter themselves. Sales teams hear objections from customers long before those patterns appear in a quarterly report.

Yet many organisations move decisions in the opposite direction. The more consequential the decision feels, the further away it travels from the people holding the most immediate context.

That does not mean every decision belongs at the edge of the organisation. Some decisions genuinely require executive involvement. Company strategy, major capital allocation, significant organisational change and material regulatory or reputational risk belong at a different level because their consequences span the whole organisation.

But routine product, design, hiring and operational decisions should not constantly require executive permission. When they do, the problem is no longer the quality of individual decisions. The problem is the decision architecture itself.

Good operating models push decisions towards the strongest combination of information, expertise and accountability. They do not automatically push them towards hierarchy.

Senior leaders should make fewer decisions as the company grows

This sounds counterintuitive because a larger organisation obviously requires more leadership. But more leadership should not mean senior leaders personally making more decisions. As the organisation grows, the nature of leadership has to change.

At small scale, a leader can create value by making many decisions directly. At larger scale, the value increasingly comes from creating the conditions in which hundreds of other people can make good decisions.

That means establishing direction, hiring strong people, defining standards, allocating resources and resolving the genuinely cross-organisational tensions that individual teams cannot solve themselves.

It also means becoming very clear about the decisions that only senior leadership can make and deliberately stepping away from the ones that do not require that level of authority.

A senior leader appearing in every important decision chain is not necessarily evidence of strong leadership. Sometimes it is evidence that the organisation has failed to scale around them. The real leverage is building an organisation capable of making thousands of good decisions without requiring that leader to be in the room.

Look at where work stops

If you want to understand how an organisation really works, do not start with the org chart. Look at where decisions stop.

Look at which decisions repeatedly move upwards. Look at which names appear in almost every approval chain. Look at the meetings that exist because nobody is completely sure who has authority. Look at which decisions are continually reopened and which teams are repeatedly waiting for permission despite supposedly owning the work.

These patterns reveal the actual operating model. An org chart tells you who reports to whom. A decision map tells you how the company really works. The difference between those two pictures can be significant.

Many organisations appear highly distributed on paper. They have executives, vice presidents, directors, heads of functions and autonomous product teams. But when you follow the flow of decisions, much of that apparent authority collapses back into a surprisingly small number of people.

Those people inevitably become overloaded. Everyone else inevitably becomes dependent. That is the opposite of scale.

The goal is not perfect decisions

There is a reason leaders find it difficult to distribute authority. People will occasionally make decisions they would not have made themselves. That is unavoidable.

When a leader has been successful by relying on strong judgement, it can be genuinely difficult to watch someone else arrive at a different conclusion. The instinct is to intervene, particularly when changing the decision appears relatively easy.

But if the standard for autonomy is that teams must always make the same decision the leader would have made, then authority has not really been distributed. People are simply being trained to imitate the person above them. The leadership question therefore needs to change.

Instead of asking, “Would I have made this exact decision?”, ask whether the decision was reasonable given the strategy, evidence, constraints and information available. Those are very different standards.

The second accepts that reasonable people can reach different conclusions. It also accepts some variation in individual decisions in exchange for dramatically increasing the organisation’s ability to act.

That is one of the fundamental trade-offs of scale. The goal is not perfect control. It is good judgement, distributed across the organisation.

Eventually, permission becomes the culture

There is a point where organisations become so accustomed to escalation that people stop noticing it.

Everyone knows which executive needs convincing. Everyone understands which meetings really matter. Everyone knows which person technically owns the decision and which person actually owns it.

People become extremely sophisticated at navigating the organisation. That sophistication can be mistaken for seniority. In reality, it often represents organisational energy being redirected away from solving the actual problem.

A company can reach the point where some of its most experienced people spend a significant part of their week preparing for approvals, managing stakeholders and predicting how decisions will be received. That is not free.

Every hour spent navigating unnecessary permission is an hour that cannot be spent understanding customers, solving problems, improving the product or developing people. Eventually, decision-making becomes part of competitive advantage.

A company where good decisions can happen quickly at the right level will learn faster than one where every important choice needs to move through an approval system.

This is why scaling is not really about how many people an organisation can add. It is about how much capability the organisation can create without adding equivalent coordination.

If every meaningful decision still requires permission, then capability has not really been distributed. It has been centralised. The organisation may be getting bigger. But it has stopped scaling.

When every decision still needs permission, you’ve stopped scaling was originally published in Bootcamp on Medium, where people are continuing the conversation by highlighting and responding to this story.