Change Through Practice

What Organisational Transformation Can Learn from Rehearsal, Performance, and Attention

Most organisational change assumes that if you explain things clearly enough, people will adjust their behaviour. New strategies are announced, values are refreshed, slide decks are shared, and everyone is expected to fall into line. When that doesn't happen, the usual diagnosis is resistance, lack of buy-in or "change fatigue".

This diagnosis is almost always wrong. The problem is rarely comprehension. The problem is that organisations are systems, and systems have immune systems.

This isn't a complaint. Immune systems are necessary. They protect against bad ideas, maintain coherence, and prevent chaos. An organisation without one would lurch from initiative to initiative, destabilised by every new proposal.

The problem is that immune systems can't distinguish between a pathogen and a transplant. They filter out novelty because it's foreign, not because it's harmful. The same mechanisms that protect the organisation from bad ideas also filter out good ones before they can prove themselves.

You are a sensor

If you work close to the actual work—with customers, with code, with the day-to-day reality of delivery—you are sensing things that people further from that work cannot see.

Most people don't disagree with this. It's in the management books. Leaders say it in town halls. And then nothing happens, because agreeing that frontline knowledge matters is easy. Building structures that actually allow it to be heard is hard.

This isn't a complaint about leadership being out of touch. It's a structural reality. Organisations are too complex for anyone to see everything. People closest to the work perceive different things than those setting the strategy. Both perspectives are partial. Both are necessary.

The question isn't whether your sensing is valid. It probably is. The question is whether your organisation has feedback loops that can hear it.

Many don't. Signals from the edges are filtered, translated, aggregated, and delayed, arriving at decision-makers as vague concerns rather than actionable intelligence. By the time leadership hears "teams are struggling with the new process," the texture of why has been lost.

This series is about what to do when you're sensing something real, but the organisation's feedback loops aren't carrying the signal. Not how to fight the system, but how to be a more effective sensor within it—and how to help the organisation develop better ways of hearing what its people are trying to tell it.

Before we go further: a diagnostic

If you're reading this, you're probably frustrated. Something you can see clearly (a better way of working, a dysfunction everyone pretends not to notice, a gap between what your organisation says and what it does) isn't changing despite your best efforts.

That frustration is real. But it's not evidence. Frustration indicates something is blocked; it doesn't indicate whether the blockage is appropriate.

Before we discuss how to amplify your signal, we need to confirm its accuracy. Some questions worth sitting with:

• Have you made your case through legitimate channels, skilfully and more than once? Or have you assumed they won't listen and skipped straight to resentment?
• Do people with more context than you, who see constraints you don't, have reasons for caution that you're dismissing as "politics"?
• Is it possible your idea isn't as good as you think, and the organisation is correctly filtering it out?
• Are you frustrated because the feedback loops are broken, or because you haven't yet developed the skills to use them effectively?

If you're honest with yourself and the answers still point to genuine signal loss (legitimate channels have been tried, the filtering is structural rather than substantive, good information is being lost), then read on.

If you're not sure, the rest of this piece may help you figure it out. But be warned: the approaches in this series have costs, and using them when you're simply wrong will damage your credibility and relationships.

How signals get filtered

Every organisation has mechanisms that filter incoming signals. This isn't dysfunction—it's necessity. Without filtering, organisations would be overwhelmed by noise, lurching from one input to another with no coherence.

The filters are familiar: governance reviews, prioritisation frameworks, alignment discussions, the polite deferral of "not now." These mechanisms exist for legitimate reasons. They create stability and focus. They also reliably filter out signals that don't fit existing mental models, even when those signals are accurate.

The key insight is that filtering is usually structural, not personal. The director who doesn't act on your observation isn't necessarily threatened or stupid. They're responding to incentives, constraints, and competing signals you may not see. Understanding this matters because it changes your approach: you're not trying to defeat enemies, you're trying to get a valid signal through a system that's working as designed but may be filtering too aggressively.

But here's the uncomfortable corollary: if you're a middle manager or team lead, you are also part of someone else's filtering. The same dynamics that frustrate you from below are dynamics you enact from above. Before you work on amplifying your own signals, it's worth asking whether you're attenuating signals from others.

What Boal can teach us

The Brazilian theatre practitioner Augusto Boal developed a set of techniques known as the Theatre of the Oppressed. His insight was that people change through rehearsal, not lecture. Instead of telling audiences what to think, he created spaces where they could pause situations, replay them, try different responses, and experience the consequences in low-stakes settings.

Boal's core methodology translates directly to organisational change. Behaviour precedes belief. People shift their minds after they experience competence, safety, and momentum, not before. New ways of working need to be practised until they feel ordinary. You don't convince people to change; you let them try it repeatedly until it no longer feels like change.

What doesn't translate is the moral context.

Boal developed these techniques for genuinely oppressed people: Brazilian peasants, factory workers facing exploitation, and communities under political persecution. His subjects had moral licence for resistance that most of us don't have. When a favela resident rehearses responses to police violence, the ethical asymmetry is clear. When a frustrated product manager rehearses how to introduce a new practice, it isn't.

This matters because invoking Boal can feel like borrowing his moral authority, implying that we, too, are oppressed and therefore justified in any means necessary. That's flattering, but it's usually not true. Most readers of this piece are knowledge workers with options, agency, and power over others. We may be frustrated, but we're not oppressed.

So take Boal's method seriously: change happens through practice, not persuasion. But don't take his moral clarity as your own. You need to earn your ethical justification separately—and for most of us, the bar is different from what it was for Boal's original participants.

Organisations as theatre

Here's something Boal understood that most change agents don't: situations are already performances. There's no neutral ground where behaviour is authentic and unscripted. Everything is staged.

Stand-ups, town halls, planning sessions, Slack channels, OKRs, dashboards: these are all stages. They shape what gets attention, what can be questioned, and what's safe to say. They reward some signals while quietly filtering others.

Most change efforts fail because they're performing on the wrong stage, to the wrong audience, with the wrong script. They try to transmit signals through channels designed to filter out exactly that kind of signal.

Evolutionary change starts from different questions:

• Where is attention already flowing?
• Which signals are being amplified without being named?
• What is treated as serious, and what is treated as noise?
• What would have to shift for different signals to get through?

You don't change a system by shouting louder. You change it by understanding how signals flow and finding the paths where your signal can actually be heard.

Safe-to-fail experiments

Here's the core practice that enables evolutionary change: small experiments that generate evidence.

Many organisations have a structural catch-22: approval requires evidence, but evidence requires doing the thing, which requires approval. If you wait for permission to experiment, you'll wait forever.

The alternative is to run experiments that are:

Small enough to be within your authority. You probably don't need permission to change how you run your own meetings, how your team communicates, or how you approach your own work. Start there.

Safe to fail. The experiment should be designed so that if it doesn't work, the cost is low and the learning is high. This isn't about avoiding failure—it's about making failure informative rather than catastrophic.

Visible enough to generate signal. An experiment nobody sees generates no evidence. The goal is to create something that others can observe, ask questions about, and potentially adopt.

Honest about what it is. Call it an experiment. Call it a pilot. Call it a bet. Call it "trying something different." You're not hiding your intent—you're being clear that this is a test, that you're learning, and that you'll share what you discover.

This is different from asking for permission to transform how the organisation works. You're not proposing a change; you're generating evidence. The proposal will follow if the experiment produces results worth proposing.

Some people will call this "going around the process." But the process probably doesn't have a path for "I'd like to try something small and see what happens." You're not circumventing legitimate oversight. You're filling a gap most organisations face: a way to learn through doing rather than through planning alone.

Rehearsal over rollout

People change their behaviour after they experience the new way, not after they agree to it. This is why training programmes fail, and pilots succeed. This is why mandates create compliance, and practice creates culture.

New rituals, new language, new habits: these need to be practised until they're ordinary. Not performed once in a workshop and then forgotten. Repeated, in context, with real stakes, until the new way is simply how things are done.

This is where the theatrical insight becomes practical. You don't need permission to start a new kind of meeting with your own team. You don't need buy-in to model a different way of giving feedback. You don't need approval to ask different questions in a planning session.

Start small. Repeat until fluent. Let others observe competence and ask questions. Expand through invitation, not instruction.

The goal isn't to sneak change past gatekeepers. It's to create a lived experience that generates a clearer signal than any proposal ever could. When someone sees a practice working, that's a different kind of evidence than a slide deck explaining why it should work.

The ethics of evolutionary change

There's a line between evolutionary change and manipulation, and it matters.

Evolutionary change is: running experiments within your authority, generating evidence through practice, sharing what you learn, inviting others to try, and building toward a proposal backed by experience rather than just argument.

Manipulation is: hiding your intent, deceiving colleagues about what you're doing, engineering outcomes while pretending to be neutral, and treating people as obstacles rather than participants.

The difference isn't always obvious in the moment. Here are some tests:

The daylight test. If someone asked you directly what you were doing and why, could you explain it honestly without embarrassment? If yes, you're probably on solid ground. If you need to obscure or spin, reconsider.

The respect test. Are you treating colleagues as people who might have legitimate perspectives, or as obstacles to be managed? Evolutionary change works with people's capacity to learn and adapt. Manipulation works around their agency.

The reversibility test. If your experiment fails or people object, can you stop? Are you creating pressure that's hard to undo? The ethics of small experiments depend partly on them actually being small.

The who-benefits test. Is this ultimately about improving how the organisation works, or about proving you were right? Ego is a powerful motivator, and it can dress itself up as a principle.

If your interventions humiliate colleagues, expose individuals, or signal your superiority, they will fail. Not just for moral reasons, but for structural ones. People will protect their dignity even when they won't protect a bad system. If you make them choose between their face and your change, they'll choose their face every time.

The risks you're taking

Even with good intent and ethical practice, evolutionary change has risks:

Your credibility. If experiments fail publicly, or if people perceive you as "going around the process," your reputation may suffer. This is true even when you're right.

Your relationships. Change creates discomfort. Some colleagues will appreciate your initiative; others will resent it. You may strain relationships that matter to you.

Your development. Evolutionary change can become a substitute for developing legitimate influence. If you always route around obstacles, you may never learn to address them directly.

Your judgement. The more you see signals that others miss, the easier it becomes to assume you're always right. You may lose the ability to recognise when you're the one who's wrong.

These risks are real. They're not reasons to never act, but they're reasons to act carefully and to stay honest with yourself about what you're doing and why.

The sensing organisation

Organisations that thrive in complex environments need to sense and respond to change. They need feedback loops that carry accurate signals from the edges to the centre and back again. They need people throughout the system who notice things, try things, and share what they learn.

This is sometimes called a "sensing organisation" or an "evolutionary organisation." The idea comes from complexity thinking: in unpredictable environments, the organisations that survive are those that can adapt continuously rather than plan perfectly.

If you're reading this series, you're probably already acting as a sensor. You're noticing things. You're frustrated that your observations aren't being heard. You want to take action.

This series is about how to do that more effectively. Not by fighting the organisation, but by being a better sensor—and by helping the organisation develop better feedback loops so that what you're sensing can actually inform what it does.

The disappearing act

The most effective change agents aren't in it for the credit. They're in it for the change. That's not selflessness as virtue—it's practical. Change that's tied to your personal brand dies when you leave. Change that's embedded in how people work outlasts you.

Months later, someone says, "Why didn't we always do it this way?" and nobody remembers whose idea it was. That's success. Not revolution. Not recognition. Just a different way of working that's become ordinary.

If you need the change to be associated with you, you'll surface too early, signal too much, and trade lasting change for temporary recognition. The quiet outcome is the real outcome.

This is the first in a series on evolutionary change. The next piece explores how to know whether your frustration is signal or noise—and what to do with the answer.