Given that this blog started off being about Kanban, before pivoting to Strategy Deployment, its mabe surprising that I haven’t written much about Deming beyond the occasional passing reference. Partly inspired by John Willis’ Profound Podcast, I’ve finally got round to do something about that. Some readers will know W. Edwards Deming and his System of Profound Knowledge well. Others may only half-recognise the name.
Deming was an American statistician and management thinker who, in the years after the Second World War, was invited to Japan to help with its industrial recovery. What he taught there, including statistical methods and a whole way of thinking about quality, systems, and the work of management, took root far more deeply than it ever had at home. Japanese industry embraced it, the Deming Prize was named in his honour, and his thinking became part of the foundation of the Toyota Production System. In time, it became what the West would relabel as Lean.
He arrived at the System of Profound Knowledge late in his life, distilling decades of work into a single framework. It is not a method or a set of practices. It is a theory. Four interrelated lenses through which to understand and transform an organisation: appreciation of a system, knowledge of variation, theory of knowledge, and psychology.
The Theory of Profound Knowledge and Strategy Deployment
All of this brings me to Strategy Deployment. It would be tempting to treat Deming’s theory as simply a useful lens to hold up against it. In other words, an analogy and an outside perspective. But the relationship runs deeper than that. Strategy Deployment is my preferred name for Hoshin Kanri, the practice of policy deployment. It grew up in 1960s Japan, with Toyota among its practitioners, as part of the wave of quality thinking that Deming had helped set in motion. At its heart, it is the Plan-Do-Study-Act cycle applied at the scale of a whole enterprise.
The X-Matrix I use to support it is not from Toyota. It is a later Western codification. But the practice it serves carries that Lean lineage. So, setting Strategy Deployment beside the System of Profound Knowledge is not introducing two strangers who happen to agree. There is a family resemblance, and there is a reason for it.
That changes the nature of the exercise from the usual posts on Strategy Deployment and other approaches. Most of those posts, such as the ones on Flight Levels and Team Topologies, are about practices that sit alongside Strategy Deployment and complement it. The System of Profound Knowledge doesn’t sit alongside anything. It sits underneath. It is less a partner than a foundation, which is exactly what we should expect of a theory that the practice descends from.
The interesting question, then, is not whether they cohere, but how deeply. And what reading Strategy Deployment through Deming’s four lenses tells us about deploying it well. Each lens, it turns out, also names a way that a deployment can drift from its own roots.
Appreciation for a System
The first lens of the System of Profound Knowledge is appreciation for a system. A system, in Deming’s definition, is a network of interdependent components that work together towards a shared aim; without an aim, there is no system. What follows is less obvious than it first sounds. The performance of any one component should be judged by its contribution to the aim of the whole, never in isolation. In fact, a component may sometimes need to run at a loss to itself so that the system as a whole can win.
Deming pressed this point hard because most management does the opposite. It manages the parts separately, sets them competing through targets and rankings, and calls the sum of those local optimisations a strategy. The result, he argued, is reliable sub-optimisation: improve the parts in isolation, and you will almost always degrade the whole.
This is where Strategy Deployment, and specifically the X-Matrix, starts. True North and Aspirations describe an aim for the whole organisation, not a stack of departmental targets. Catchball connects the parts up, down, and across so that the strategy is shaped by the whole rather than imposed on it. And the X-Matrix also makes the system visible. It isn’t a list of goals. It is a set of elements held together by correlations. Coherence lives in the relationships, not in cells.
So the warning for deployment is to resist letting the X-Matrix decay into a scorecard. Each team should not own its own column and chase its own number. The moment that happens, you have rebuilt exactly the sub-optimisation Deming cautioned against. The discipline is to keep attention on the correlations. That is how the parts contribute to one another, rather than on the parts alone.
Knowledge about Variation
The second lens of the System of Profound Knowledge is knowledge about variation. Everything varies. Every process, every measure, every result. Deming, building on the work of Walter Shewhart, taught that the central skill is telling two kinds of variation apart. Common-cause variation is the ordinary fluctuation of a stable system, and the noise it produces simply by running. Special-cause variation is something specific and assignable, a genuine signal that something has changed. The danger lies in confusing the two. React to common-cause noise as though it were a signal, or what Deming called tampering, and you will reliably make a stable system worse rather than better.
Deming thought this mattered enormously because managers who cannot read variation end up blaming people for the behaviour of the system. They chase every wobble in the numbers and set targets the data could never have supported. Most of what goes wrong, he reckoned, belongs to the system rather than to the people working within it.
Strategy Deployment has a place for this in the Evidence dimension. Evidence is how we learn whether our strategies are doing anything. But evidence varies, and a single reading moving up or down is not necessarily telling us anything at all. I’ve written before that X-Matrix correlations describe association, not causation. Variation adds the other half of the caution. Even a real association can be buried under noise, and a number that twitches is not the same as a number that has changed.
So the implementation lesson is not to tamper. When the Evidence wobbles, the question is whether the system is talking or a special cause has appeared. This should be asked before abandoning a strategy or doubling down on it. This is another argument for Continuous Strategy. A regular cadence lets you watch the trend rather than chase the point, which is the difference between reading variation and being read by it.
Theory of Knowledge
The third lens of the System of Profound Knowledge is the theory of knowledge. Deming insisted that we be clear about how we know what we think we know. He held that there is no knowledge without theory, and that, in his words, “management is prediction”: any plan worth the name is a prediction about what we think will happen if we act in a certain way. And we only learn when we compare what actually happens against what we predicted. Experience on its own teaches nothing. It is the comparison of outcome to expectation that turns it into knowledge, which is the real work of the Plan-Do-Study-Act cycle.
Deming cared about this because so much of management substitutes imitation and assertion for it. Copying another organisation’s practices without understanding the theory of why they worked. Or declaring success after the fact by quietly rewriting what had been expected.
Strategy Deployment treats strategy as a set of hypotheses rather than a set of certainties. As I put it when measuring the X-Matrix, we believe the Aspirations are achievable, we believe the Strategies will help, and we believe the Evidence will show progress. And then we test those beliefs. The order of the X-Matrix is the same instinct made structural: working through Aspirations, Strategies, and Evidence before settling on Tactics, so that the Tactics are tested against what we expect to see rather than justified after the fact. Deming would recognise the avoidance of retrospective coherence as a defence against fooling ourselves.
So when deploying, write Strategies and Tactics as predictions you could be wrong about, and decide in advance what Evidence would change your mind. A strategy that can’t be wrong isn’t empirical. It’s a belief in fancy dress. It is also worth remembering, as I argued in The Experiment Trap, that prediction is not the only way to learn. Sometimes the honest move is to probe rather than to hypothesise, precisely because a theory of knowledge includes knowing the limits of what we currently know, or can know.
Psychology
The fourth lens of the System of Profound Knowledge is psychology. An understanding of people, of what motivates them, and of how they respond to the way they are managed. Deming drew a sharp line between intrinsic motivation (the pride, curiosity, and joy in work that people are born with) and extrinsic motivation (the rewards and punishments applied from outside). His claim, which unsettled many managers, was that the standard apparatus of motivation through incentive pay, ranking, grading, or management by the numbers, does not add to intrinsic motivation but crushes it, and breeds fear in its place.
Deming put “drive out fear” among his famous fourteen points for a reason: fear makes people hide problems and feed their managers comfortable figures, and stop cooperating, which quietly corrodes every other part of the system.
Strategy Deployment’s whole stance of alignment with autonomy is psychological at root. Catchball only works if people feel able to disagree. Under fear, it collapses into compliance dressed as consent. Engaging the people closest to the problem, or closest to the context, treats strategic contribution as everyone’s work. It should be a source of pride, rather than an overhead handed down from above.
Which makes catchball a canary. If catchball produces nothing but agreement, the psychology is wrong. Fear or extrinsic pressure is in the room, and the deployment will be hollow, however elegant the X-Matrix looks. The thing to watch for is not whether people nod, but whether they feel safe enough to push back.
A System of Lenses
Deming was insistent that the four parts of the System of Profound Knowledge cannot be taken in isolation. You can’t appreciate a system without understanding its variation. And you can’t have a theory of knowledge that ignores the psychology of the people doing the knowing. The four are themselves a system.
I find that rather pleasing, because it is precisely how the X-Matrix works. The X-Matrix is not a list of goals. Its meaning lives in the correlations between its elements rather than in any element on its own. What I have described using the term messy coherence. Deming’s theory has the same character: a way of seeing systems that is itself a system, its parts making sense only in relation to one another. The tool we deploy strategy with, and the theory that underwrites it, are built on the same idea.
So the System of Profound Knowledge isn’t a rival to Strategy Deployment, and it isn’t quite a complement either. It is closer to the theory in Strategy Deployment’s ancestry. The thinking the practice grew out of, carried forward, and made into a way of working. That is why the coherence runs so deep: it is inheritance, not coincidence. And reading Strategy Deployment through the four lenses does more than confirm it. It hands us a short list of the ways a deployment drifts from its lineage: sub-optimising the system, tampering with variation, mistaking conviction for empiricism, and neglecting the psychology that makes any of it possible. The coherence is reassuring. The warnings are arguably more useful.