Levels and Leverage: Stop Fixing Symptoms in Organisations

The Hidden Constraint

A friend of mine runs a mid-sized agency. Good people, decent product, clients who keep coming back. But the place had developed what everyone called a “silo problem.” Teams were not talking to each other. Handoffs were messy. Blame was the default currency when anything went wrong. People were polite in meetings and brutal in private Slack channels.

So the leadership team did the obvious thing. They launched a culture initiative. New values were workshopped, printed, and mounted on the breakroom wall. There was a company-wide town hall where the CEO used the word “collaboration” so many times it lost all meaning. An email went out about “who we are becoming.” A team-building offsite was booked. Everyone went kayaking.

Nothing changed.

Not because the people lacked goodwill. Not because the kayaking was bad. But because every bonus in the company still rewarded individual output. Every promotion still went to the person who took credit most visibly. The reporting lines still meant that helping another team was, in practical terms, time stolen from the work you were actually measured on. The words on the breakroom wall said “collaboration.” The pay structure said “look after yourself.”

People are not stupid. When what gets said and what gets rewarded are two different things, people follow the money. They always follow the money. And every time leadership launched another values campaign that did not touch the incentive structure, something worse than apathy grew: cynicism. People noticed the gap between the poster and the payslip. They noticed that the “culture initiative” had changed the wallpaper while leaving the plumbing exactly as it was. And each round of this eroded trust a little further. Not because people were negative. Because people were paying attention.

This is a post about a deceptively simple idea that changes how you see organisations once you grasp it: problems exist at different levels, and most failed interventions fail not because the approach was bad, but because it was applied at the wrong level. The technique and the problem did not match. The boundary leadership drew around “the issue” was too narrow, or too abstract, or excluded the very structure that was generating the pattern in the first place.

The Levels People Actually Work Across

Every organisation — every team, every family, every system involving people — operates across distinct levels. These are not boxes on an org chart. They are layers of cause and effect, each one shaping the behaviour of the people inside it. Understanding these levels is the difference between an intervention that actually changes how people work together and one that makes leadership feel productive while the original problem quietly regenerates itself.

Think of it as a stack, running from the most concrete and visible to the most abstract:

1. What actually happens at the edges. This is where your organisation meets the real world: the customer interaction, the frontline moment, the point of service delivery. This is where the system’s output gets experienced by real people. You rarely fix things at this level directly — it is the layer that reflects everything above it. If your customer experience is poor, the answer is almost never to retrain the receptionist. It is somewhere upstream.
2. How the work gets done. The processes, tools, workflows, and standard operating procedures that shape people’s daily experience. This layer determines whether someone can actually do their job well or whether the system makes competence nearly impossible. It is concrete, observable, and often the first place leaders look when something goes wrong — which is fair when the problem is genuinely procedural, and misleading when it is not.
3. What gets measured, rewarded, and punished. This is the gravitational field. People are rational. If the incentive structure rewards individual throughput, collaboration will not emerge no matter how many team-building days you run. If the KPIs measure speed, quality will be sacrificed. This is where most leadership interventions should land and where they most often do not — because changing incentives means admitting the current ones were wrong, and that conversation is uncomfortable.
4. Who enters the system and what they are given. Hiring and training. This layer determines the raw material the organisation works with. It is slower than changing a process or a bonus structure — you cannot retrain a workforce overnight — but it compounds powerfully over time. The wrong hire at a senior level can corrupt an entire team’s incentive layer. The right development programme can make a broken process survivable until it gets fixed.
5. What is actually tolerated. The unwritten rules. Not what the policy says, but what really happens when someone makes a mistake, disagrees with a senior person, or raises a problem nobody wants to hear about. Culture is not what leadership declares in a town hall. It is what the organisation tolerates on a Tuesday afternoon when nobody important is watching. This layer is shaped by everything below it and constrains everything above it. You cannot install culture. You can only create conditions from which it emerges.
6. The story leadership tells. The narrative about what the organisation is, where it is going, and why any of it matters. This is the most abstract level — and, paradoxically, both the most powerful and the most impotent. Powerful because a coherent narrative aligns attention and meaning. Impotent because a narrative that contradicts the incentive and cultural layers is not just ineffective, it is corrosive. It creates exactly the cynicism I described at the start.

Here is the critical insight: higher levels constrain the ones below them, but lower levels reveal the ones above them. If frontline reality is broken, the driver is usually two or three levels up. If the leadership narrative is inspiring but nothing is changing on the ground, the blockage is at a level below the narrative. Diagnosis means tracing the problem to the right level before choosing the intervention.

Stop fighting the moment. Redesign the loop. The question is not whether your intervention was strong enough. The question is whether it was aimed at the right floor of the building.

Why People Keep Fixing the Wrong Thing

Once you see these levels, you start to see why so many well-intentioned efforts fall flat. Five patterns show up again and again.

1. Talking when the problem is structural

This is the most common failure in organisations. Leadership identifies a behavioural issue — people are not collaborating, or quality is slipping, or initiative has dried up — and responds with communication. Emails. Town halls. Values statements. But communication operates at the narrative level. Behaviour is driven by the incentive level. The intervention is three floors above the driver. It is the organisational equivalent of trying to fix a leaking pipe by writing a strongly worded memo about water conservation.

2. Drawing the boundary too tightly

A team fixes its internal process beautifully — new workflows, better documentation, cleaner handoffs within the team. But the real constraint was at the interface between that team and the one upstream, where incomplete inputs kept arriving. The process improvement worked perfectly within the boundary they drew. It changed nothing in the actual system, because the problem was at the seam between two groups of people, not within either one. The boundary was too tight.

3. Local wins, system losses

Each department optimises for its own numbers, which were set without considering how departments depend on each other. Sales optimises for new revenue and promises delivery timelines that Operations cannot meet, because Operations is optimising for cost containment. Each team is performing well by its own metrics. The system is failing. And the people caught in the middle — the project managers, the account leads, the frontline staff who have to look the client in the eye — absorb the stress of a structural conflict that has nothing to do with their competence.

4. The metric becomes the goal

A KPI is introduced as a proxy for an outcome that matters. Over time, the proxy becomes the target. People begin optimising for the number rather than for the thing the number was supposed to measure. Call centres measuring “average handle time” train staff to get callers off the phone as fast as possible rather than to actually solve their problems. The metric improves. The experience degrades. And the people working the phones know exactly what is happening — they just cannot say it, because the metric is what their performance review is built on.

5. The hierarchy flips upside down

In a healthy organisation, the hierarchy exists to support the people doing the work. Leadership removes obstacles, secures resources, and protects the conditions under which good work can happen. In an inverted hierarchy, the frontline exists to serve leadership’s narrative. Data is curated upward to confirm the strategy. Bad news gets filtered. Reality at the edges becomes invisible to the top floor, and the top floor makes decisions based on a version of reality that no longer exists. The person who tells the truth is seen as “not a team player.” The person who tells leadership what it wants to hear gets promoted. The hierarchy, which should be a support structure, becomes a reality-distortion field.

Every one of these failures has the same root cause: the intervention was aimed at the wrong level, the wrong boundary, or both. The people involved were not incompetent. They were solving the right problem at the wrong address.

Three Levers That Actually Move Things

If wrong-level interventions are the disease, what does it look like to get it right? Three principles consistently separate organisations that actually change from organisations that just talk about changing.

1. Map the boundary honestly

Before you choose an intervention, draw the boundary of the problem. Not the boundary that is comfortable. The boundary that is honest.

Most diagnostic errors happen because the boundary is drawn too tightly. The problem is framed as belonging to one team, one process, one individual — when it actually spans multiple boundaries and involves the dynamics between people who may not even realise they are part of the same system.

The question is literal: What has to be inside the fence for this problem to be solvable? If you are solving a quality problem but the fence excludes the staffing model, you will optimise inside a constraint that is itself the problem. If you are solving a strategy execution problem but the fence excludes the bonus structure, you will run alignment workshops that do not survive contact with the next pay cycle.

Draw the boundary wide enough to include every level that touches the outcome. Then — and only then — figure out which level inside that boundary is doing most of the work.

2. Start at the lowest level that will shift the outcome

Once you have identified the right level, intervene at the lowest concrete point that changes what actually happens. Do not change the story if changing the incentive will do. Do not retrain people if fixing the process will do. Do not redesign the process if filling three vacant roles will do.

This is organisational minimum necessary force. Higher-level interventions — narratives, culture programmes, values refreshes — are slower, vaguer, and harder to verify. Lower-level interventions — changing what gets measured, fixing the workflow, hiring the people you need — are concrete, measurable, and fast to validate. Change the incentive and you will know within one performance cycle whether behaviour shifted. Change the narrative and you may wait two years before admitting it did not work — by which time you have compounded cynicism and burned the organisation’s appetite for any kind of change.

This does not mean narrative and culture are unimportant. It means they are emergent. They emerge from the structural layers below. Get the incentives right, the process right, the hiring right, and culture follows. Try to install culture directly, without touching the structure beneath it, and you get a poster on the wall that nobody believes.

3. Make the hierarchy serve the people doing the work

In a functional organisation, each layer of the hierarchy has a specific job in relation to the people around it. When these relationships are healthy, the system adapts. When they invert, the system becomes rigid and self-deceiving. Four things distinguish the healthy pattern:

• Truth travels upward without distortion. The primary function of senior leadership is to create conditions where honest information — especially bad news — reaches decision-makers at the same speed as good news. This means the person who raises an uncomfortable truth is treated as a resource, not a problem. The moment leaders begin curating the information they receive to confirm a strategy they have already committed to, the hierarchy has flipped. The system is now optimising for comfort rather than reality.
• The middle layer absorbs volatility. The function of middle management is to shield the people doing the work from the constant switching costs of strategic change. Not to block information, but to buffer the disruption. Every strategic pivot imposes a transition cost on the humans who have to implement it. Middle management’s job is to absorb that cost — to translate strategic direction into operational reality at a pace people can actually metabolise. When middle management is bypassed or hollowed out, strategic directives hit the frontline unfiltered and you get the initiative fatigue described above.
• Numbers inform decisions; they do not replace them. Metrics are inputs. The moment a metric becomes the decision — “hit the number regardless of context” — it stops measuring reality and starts distorting it. People are smart. They will find ways to hit any number you set. The question is whether hitting the number also produces the outcome you actually care about, or whether the number and the outcome have quietly divorced.
• Strategy respects what people can actually deliver. A strategy that ignores the capacity, skill, and structural constraints of the people who have to execute it is not a strategy. It is a wish. Brilliant strategic thinking that requires capabilities the organisation does not have, cannot build in the available time, and has not resourced is an exercise in aspiration dressed as leadership. The people on the ground know this. They just cannot say it out loud without being labelled “resistant to change.”

Hierarchy is not the problem. Inverted hierarchy is the problem. The question is not whether to have levels. It is whether those levels are serving the people doing the work, or serving themselves.

The Leverage Ladder Memo

Ideas are free. What costs something is the discipline to apply them before you have already committed to the wrong intervention. What follows is a structured way to diagnose the right level before you spend the budget, book the workshop, or send the email.

What Healthy Boundaries Look Like Between People

So far, we have been talking about levels of intervention. But boundaries matter just as much between people and between groups as they do between layers of the stack. And in practice, most of the pain in organisations lives at the boundaries — the handoffs, the interfaces, the spaces between teams where nobody is quite sure whose job it is and everybody is quite sure it is not theirs.

Healthy boundaries between people and teams have three characteristics:

• Clarity about who owns what. Not in the sense of rigid territory-marking, but in the sense that when something goes wrong at the seam between two groups, there is an agreed process for who picks it up. Most “accountability problems” in organisations are actually boundary problems — the gap between two roles where a task falls and neither person catches it, because neither person knew it was theirs.
• A working feedback loop at the interface. The people on either side of a boundary talk to each other about what is working and what is not — regularly, specifically, and without waiting for a crisis. When this loop breaks down, small problems accumulate silently until they produce a failure that looks sudden but was weeks or months in the making.
• Permission to raise problems without penalty. The boundary between a junior employee and their manager, or between a team lead and a director, is only healthy if honest information can cross it in both directions. The moment raising a concern becomes a political risk, the boundary hardens into a wall. Information stops flowing. And leadership makes decisions in an increasingly fictional version of reality.

If you trace most organisational dysfunction back far enough, you will find a boundary where honest information stopped crossing. A handoff where nobody was minding the seam. A reporting line where truth got filtered on the way up. The structure was fine. The boundary between people was broken.

Key Takeaways

• Organisations operate across six distinct levels: what happens at the edges, how the work gets done, what gets measured and rewarded, who enters and what they are given, what is actually tolerated, and the story leadership tells. Each level constrains the ones below it. Each level reveals the ones above it.
• Most failed interventions are right problem, wrong level. The diagnosis is often accurate. The intervention is aimed at the wrong floor. Narrative solutions for incentive problems. Training solutions for capacity problems. Communication for structural failures.
• Start at the lowest level that will shift the outcome. Lower-level interventions are concrete, fast, and verifiable. Higher-level interventions are slow, vague, and prone to generating activity without impact. Change the incentive before you change the story.
• Healthy hierarchy serves the people doing the work. Truth flows upward. The middle layer absorbs volatility. Numbers inform decisions without replacing them. Strategy respects what people can actually deliver. When these relationships invert, the system becomes self-deceiving.
• Most pain lives at the boundaries between people. Unclear ownership, broken feedback loops, and penalties for honesty — these are where the real dysfunction accumulates. Fix the seams, and much of the interior fixes itself.
• Use the Leverage Ladder Memo before committing resources. Name the outcome. Walk each level. Find the driver. Design one precise intervention. Add one support. Set a review window that matches the level. One page. The discipline is in the constraint.

If the fix is at the wrong level, the problem comes back. It comes back because the structure that produced it is still intact. The symptom was suppressed. The driver was untouched. And the organisation, having invested time and credibility in the wrong move, is now less willing to try again — not because change is impossible, but because the wrong change burned the appetite for the right one.

Stop fighting the moment. Redesign the loop. Find the level where a precise shift produces the largest downstream effect — and start there.

The final post in this series takes everything — stocks and flows, feedback loops, delays, thresholds, resilience, hierarchy, boundaries — and asks the last question: what does it look like when you stop trying to control a system and start learning to move with it?