You can’t transform what you don’t understand
Why most operational excellence starts in the wrong place
The moment you realise you don’t understand the work
A few years ago, I was working with a large public service organisation that had formally launched what it called a ‘transformation programme’. On paper, everything looked right. There was a central team, a roadmap, a set of workstreams, and a growing sense of urgency from the top.
In the first session with the senior team, I asked what seemed like a straightforward question: where does demand come from, and what happens to it when it arrives?
The response was hesitant. There were fragments of answers, some confident assertions, and then disagreement. It was clear quite quickly that there wasn’t a shared understanding.
So rather than debate it, we went to look.
We spent time with frontline teams. We sat alongside people as calls came in, as cases were logged, as work moved — or didn’t move — through the system. Within a few hours, a different picture emerged. Demand was more varied than expected. A significant proportion of it existed only because earlier work had not been resolved properly. Cases moved back and forth between teams. Decisions were delayed, revisited, or escalated in ways that made sense locally but created delays overall.
What became obvious was not that the organisation lacked effort or intent. It was that it did not understand its own work well enough to improve it.
The transformation programme had not yet failed. But it was already pointed in the wrong direction, because it was built on assumptions rather than observation.
That pattern is more common than most people would like to admit. Smart people, working hard, creating plans for systems they have not really seen.
Why this matters more than most leaders realise
From a senior leadership perspective, it is entirely reasonable to begin with strategy, structure, and plans. That is how most organisations are designed to operate. Direction is set, resources are allocated, and execution is expected to follow.
The difficulty is that operational performance does not respond neatly to that model, particularly in complex environments such as public services or large organisations with multiple interdependencies.
Work rarely flows according to the organisational chart. Demand does not arrive in clean, pre-defined categories. Decisions are taken locally, often under pressure, and often with incomplete information. People adapt constantly in order to keep things moving, developing workarounds that are invisible at senior levels but essential to day-to-day functioning.
If that reality is not understood, then any attempt to improve performance risks missing the point.
This is where many operational excellence efforts begin to drift. The problem is framed as one of alignment, discipline, or capability. The response is to introduce more governance, more reporting, more initiatives. Each of these interventions may make sense in isolation, but collectively they often add weight to a system that is already struggling to function.
The underlying issue is usually structural rather than behavioural. The system itself is producing the outcomes that leaders are trying to change.
If demand is fragmented, duplication is inevitable.
If measures incentivise the wrong behaviours, distortion will follow.
If decisions are taken far from the work, delay and rework are built in.
For senior leaders, the consequence is not just underperformance. It is the gradual erosion of confidence in change itself. Significant time and resource are invested in programmes that never quite land, and the organisation becomes more cautious, not less, about trying again.
How to start from the work rather than the plan
The alternative is straightforward to describe but requires discipline to carry through: begin with understanding the work as it actually happens.
In practice, this involves three areas of focus.
The first is demand. It is not enough to know how much work is coming in; it is necessary to understand what that work consists of. What are people asking for? How predictable is it? How much of it is ‘failure demand’ — activity that exists because something has already gone wrong?
In one organisation, more than half of incoming demand fell into this category. Until that was visible, no amount of process redesign would have produced meaningful improvement.
The second is flow. Work needs to be followed from start to finish, across boundaries. Where does it wait? Where does it return for rework? Where does it depend on approvals or handoffs that introduce delay?
These points of friction are often not captured in formal reporting. They sit in the spaces between teams, in unclear ownership, and in the small decisions that accumulate over time.
The third is decision-making. It is important to understand who is making decisions, on what basis, and with what information. Are decisions taken where the knowledge sits, or are they escalated? What happens when a case does not fit the standard process?
This is where the language of cybernetics becomes practical. The organisation can be seen as a system that regulates itself through feedback. Where feedback is timely and accurate, the system can adapt effectively. Where it is delayed, distorted, or absent, the system compensates in ways that often create additional work.
Only once these elements are understood does it make sense to intervene. Even then, the aim is not to design an ideal future state in one step, but to make targeted changes and observe their effects.
Constraints can be removed. Measures can be adjusted. Decision-making can be moved closer to the work. Each intervention provides information about how the system responds.
There are predictable risks. Leaders may become impatient and attempt to scale changes too quickly. Teams may revert to designing processes rather than improving flow. Existing measures may pull behaviour back towards familiar patterns.
Maintaining focus on the work itself, rather than on abstract models, is what keeps the effort grounded.
What this approach requires in practice
This way of working does not demand a large, centralised programme, but it does require different conditions from those typically associated with transformation.
Time is the most immediate requirement. Senior leaders need to spend time engaging with the work directly. This does not mean occasional visits or presentations, but sustained attention to how the system actually operates. For many, this represents a significant shift in how they use their time.
In terms of capability, the organisation needs people who can see systems rather than isolated processes. This includes the ability to observe without jumping to conclusions, to ask questions that reveal how work is really done, and to facilitate improvement in a way that builds local ownership.
Financial investment is rarely the limiting factor. Most improvements come from removing waste, reducing rework, and improving flow. The more significant investment is in attention, consistency, and the willingness to stay with the work long enough to understand it.
Data also needs to be treated differently. Aggregate performance metrics have their place, but they do not reveal how work moves through the system. More useful are measures that reflect end-to-end flow, types of demand, and the frequency of rework.
Perhaps the most important condition, however, is permission. People need to be able to surface problems without fear of blame. They need to be able to test changes without having to guarantee success in advance. They need to be able to question existing structures and measures without being seen as resistant.
Without this, the system will tend to defend itself, and improvement efforts will stall.
What changes when you start from understanding
When organisations take this approach, the nature of change shifts in noticeable ways.
The number of initiatives typically decreases, but their impact increases. Attention is focused on a smaller set of changes that are grounded in how the system actually works.
Performance improvements become more meaningful. Rather than short-term gains against isolated targets, organisations see reductions in delay, improvements in quality, and a more stable flow of work.
Leadership becomes more effective, not because plans are more detailed, but because decisions are based on a clearer understanding of the system. There is less reliance on assumptions and more engagement with reality.
Perhaps most importantly, the experience of change for staff is different. When changes reflect the conditions in which people are working, they are more likely to be adopted and sustained. Change stops feeling like something imposed from above and becomes something that emerges from within the system.
Operational excellence, in this sense, is not primarily about tools or methodologies. It is about developing a clear view of the system that produces current results, and then improving that system in a deliberate and informed way.
Without that understanding, even well-intentioned efforts are reduced to guesswork. With it, improvement becomes both more effective and more durable.
About the Author:
Benjamin Taylor is Managing Partner of RedQuadrant, a service transformation consultancy, and chief executive of the Public Service Transformation Academy, a not-for-profit social enterprise. He’s also on the Board of SCiO (Systems and Complexity in Organisation), the systems thinking practitioner professional body, and of the Operation Excellence Society.
