Imitate Yesterday’s Winners or Shape the Market. One Path Ends in Irrelevance

This is how competitive advantage dies in a mature enterprise. Not as collapse. As latency. A CEO will not sign a check for latency. There is no line item for it. There is no capital request titled “purchase a slower future.” What gets funded is insight. What gets funded is tooling. What gets funded is the illusion that the company is becoming more intelligent because it can describe itself more clearly. The problem is that description is not control. Description is the rearview mirror. Control is the ability to intervene. Control is the ability to make the next outcome different from the last one, on purpose, without convening the entire company to grant permission. That is the dividing line between an enterprise that imitates and an enterprise that shapes. If you want a clean way to see the difference, use a test that executives instinctively understand. Look five years out and ask a blunt question. Whose name is going to be on the assets you operate. Not your logo. Your assets. Your plants, your network, your margin engine, your customer relationships, your distribution footprint. The stuff that defines your power. Whose name will it wear. Because in a world that moves at high dynamism, the market does not acquire laggards because they are stupid. The market acquires laggards because they are late. They are late to see the shift, late to commit, late to reconfigure, late to learn, late to stabilize the new reality into a repeatable operating state. They become a value play. Their competitors become the narrative. This is not a morality tale about boldness. It is a systems tale about how firms learn. It is about whether your operating model is forward facing or rear facing. Most companies still run their strategy from the rearview mirror. They call it best practice. They call it benchmarking. They call it learning from leaders. They call it prudence. What they are actually doing is outsourcing their future to yesterday’s correlates, then wondering why they never arrive in time.

The Imitation Trap

There is a seductive promise inside modern management culture. It says that the winners have been identified. Their traits can be observed. Their practices can be documented. Their metrics can be benchmarked. If you adopt what they adopt, you can become what they are. That promise used to be partially true, or at least true enough to feel real. When dynamism was lower, the system had more time to correct. When product complexity grew more slowly, the enterprise could keep up through human inference. When the environment moved at a speed compatible with governance gates, the organization could afford to learn late and still be competitive.

Today, imitation is not a strategy. It is a delay tactic. Because what you can observe about winners is overwhelmingly correlated. It is surface visible. It is the output of their system, not the system itself. You can see their portfolio moves, their operating cadence, their investment patterns, their talent density, their performance. You cannot see the mechanism that converts signal into action inside their enterprise without friction. You cannot see the permission architecture that lets decisions land early and stay landed. You cannot see the decision logic that is practiced daily until it becomes muscle memory. So imitation reliably produces a particular outcome. It makes you look modern while you remain structurally late. That is why correlation based benchmarking quietly trains enterprises to lose. It teaches them to chase symptoms of advantage instead of building the mechanism that generates advantage. It teaches them to copy the visible play, not the operating code. It turns a company into a follower with excellent analysis. In a slow world, a follower can survive. In a fast world, a follower becomes inventory. This is the piece executives miss because they have spent their careers winning with inference. They are intelligent. They are experienced. They can walk into a messy situation and diagnose what is happening faster than most people can name it. That is a real strength. It is also the trap. They assume the enterprise can scale what they can do as individuals. It cannot. Human inference does not scale across a modern enterprise at modern speed. It bottlenecks at the very moment it matters. When uncertainty rises, the organization does not speed up its inference. It slows down its commitments. It adds gates. It adds reviews. It adds escalation. It adds interpretation layers. It adds committees. It adds what feels like control and is actually delay. That delay becomes your speed limit.

The Speed Limit Inside Every Enterprise

Every enterprise has a learning rate limit. It is the maximum speed at which the organization can sense, interpret, decide, and act while staying coherent and stable. When the environment moves slower than that limit, you can govern through humans. You can rely on escalation. You can settle truth in meetings. You can run the company through approval gates and still get away with it. When the environment moves faster than that limit, the enterprise has only two options. It can redesign how it learns and decides so that it can keep up. Or it can protect itself by slowing down, then calling the slowdown “governance.”

Most choose the second option because it feels responsible. It looks like discipline. It creates the appearance of risk management. It also ensures that the enterprise experiences the future as a surprise. This is why you cannot drive an enterprise in the rearview mirror. Not because history is useless. History is valuable. It is evidence. The problem is that correlation and hindsight tell you what happened. They do not tell you what to do next. They do not tell you what will happen if you intervene. They do not tell you which lever is causal and which is cosmetic. They do not tell you which improvement will compound and which will create hidden second order cost. That distinction is the entire game. If you look at your own causal maps, the ones you built to explain manufacturing productivity stagnation and decline, the structure is the message. There are triggers and controls at the top. There are operational challenges. There are contributors that create the stagnation state. There is a target node. There are downstream outcomes and secondary challenges that feed back into the system. The map is not saying “do more of the usual.” It is saying that productivity is a complex system outcome. It is an accumulated advantage loop when managed correctly and an accumulated penalty loop when managed late. Correlation can describe that loop after it plays out. Causality is what lets you interrupt it while the cost is still small. The reason this matters is painfully simple. Most of the damage in operations is not binary. It is area under a curve. A drift condition that persists for 30 days is not 15 days worse than one that persists for 15 days. It is often multiples worse because it compounds into scrap, rework, premium freight, schedule instability, customer concessions, morale decay, and the quiet reallocation of your best people into firefighting. The economics of delay are nonlinear. So when a CEO asks “why did we miss this,” the honest answer is often “we saw it.” The enterprise saw it. Someone knew. Multiple people knew. But the organization could not convert what it saw into action without climbing a permission staircase. And the staircase is engineered to slow action until it is safe. Safe means consensus. Consensus means time. Time means cost. That is the rearview mirror enterprise. A forward-facing enterprise designs the opposite. It designs for early interventions, bounded by clear guardrails, so that learning happens while the cost is still cheap. It makes the organization capable of moving, not merely capable of explaining.

Correlation Gives You Patterns. Causality Gives You Leverage

A correlation economy is built on the belief that if you can observe enough, measure enough, and compare enough, you can become what the leaders are. It produces league tables. It produces maturity models. It produces lists of “moves.” It produces confidence in the narrative because the narrative is backed by data. The problem is that data about what is associated with performance is not the same thing as a mechanism for producing performance. Causality is a different species of capability. It is not primarily about better analytics. It is about a different relationship between the enterprise and reality. It says that the firm must be able to answer three questions repeatedly, in the flow of work. What do we believe causes what. What will happen if we intervene. What evidence will tell us whether we were right. That is forward facing. It is also uncomfortable, because it forces commitments. You cannot hide behind “leaders tend to do X.” You have to say “in our system, if we change Y, outcome Z should move, and here is why.” When you are wrong, you do not debate. You update the model. The enterprise learns. That is how shaping enterprises compound advantage. They do not merely have better strategies. They have better feedback loops. This is also where the application of the 20.20.60 model becomes an indictment of how most consulting shaped transformation is sold. The first twenty is choosing the right things. That is where benchmarking loves to live. It helps you name priorities. It helps you set targets. It helps you adopt the vocabulary of winners. It creates an external reference point for what “good” looks like. The second twenty is doing those things right. That is competence, discipline, execution quality. It is the ability to translate intent into reliable operations. Many firms never truly reach it because they treat execution as willpower, not as architecture. The last sixty is focus. It is staying pointed at the right work long enough for compounding to occur. It is the ability to protect attention from the enterprise’s own interruptions. Correlation based programs barely touch the sixty, because they do not change the mechanism that generates interruption. They leave decision gates in place. They leave escalation as the path to action. They leave human inference as the regulator. So the enterprise remains fragile. It remains distractible. It remains late. It remains dependent on heroes.

The result is predictable. Leaders get busier and outcomes become less controllable. The organization feels like it is running harder to stay in place. Causal operating models do the opposite. They reduce the need for human arbitration by encoding decision logic into runtime structures. They collapse the distance between signal and action by designing permission, not by asking people to be braver. They make improvement a capability, not an event. This is what a shaping enterprise actually means. It means the enterprise can create outcomes deliberately, then repeat them.

The Reader Choice

If you want to see the mainstream correlation framing in its cleanest and most influential modern form, read this report straight through, then come back and ask yourself what it actually equips a CEO to do on Monday morning when the system is drifting. https://www.mckinsey.com/mgi/our-research/the-power-of-one-how-standout-firms-grownational-productivity Do not read it as a bad argument. Read it as a complete argument for a different era. Read it as the kind of work that flourishes when the enterprise can still be run by inference and gates because the world moves slowly enough to allow it. Read it as rearview mirror excellence. Then ask the forward-facing question. What operating mechanism does it give you that changes how your company delivers and shapes outcomes. That is the gap. That gap is the story.

Why Rearview Mirror Excellence Leads to Irrelevance

The most dangerous thing about a rearview mirror strategy is not that it is wrong. It is that it is often right enough to feel righteous. Correlation tells you that a small number of standout firms drive a disproportionate share of productivity gains. That is a legitimate observation. It is also a story that can be weaponized into the wrong executive conclusion. It can tempt leaders into believing that the answer is simply to copy the “moves” that standouts made, to get bolder, to be more aggressive, to scale what works. In a boardroom, that sounds like strategy. In an enterprise, it often becomes theatre. Because the constraint is not whether you know what bold looks like. The constraint is whether your organization can absorb boldness without becoming unstable. The constraint is whether

your decision system can make and keep commitments early enough to capture the advantage. The constraint is whether your permission architecture can translate insight into action without climbing a staircase that is designed to slow you down. Rearview mirror excellence routinely ignores that constraint because it treats the enterprise as if it can execute the same move set as the winners, as long as leadership chooses to. But leadership choice is not the problem. Architecture is. A shaping enterprise is not the one that declares boldness. It is the one that can run boldness as a repeatable operating pattern, with bounded risk, and fast learning. That is not a mindset. That is an engineered loop. This is why, in high dynamism, the winners stop looking like “the most efficient companies” and start looking like the most controlled companies. Controlled not in the bureaucratic sense. Controlled in the cybernetic sense. They can regulate themselves against disturbance because their regulator has enough variety and speed to match the environment. They do not need an empire of meetings to settle reality. They have built a system that can decide, act, and learn at the edge. This is what makes them market shaping. They do not simply respond to the market. They shape what becomes normal, because they can move earlier and stabilize earlier. Everybody else imitates later. Then they explain later. Then they get acquired.

The Enterprise That Can Only Fix Things Once

There is a pattern you have named that deserves to be said bluntly, because CEOs recognize it instantly when it is finally spoken in plain language. Most enterprises can only fix things once. They can run a great program. They can rally talent. They can execute a transformation sprint. They can clean up a plant. They can reset a supply chain. They can compress costs. They can create a year of heroism that produces visible results. Then the gains fade. Not because people stopped caring. Not because the leaders were insincere. Because the underlying system of decision making and permission did not change. The organization reverted to its default speed limit. The same gates returned. The same escalations returned. The same interruptions returned. The enterprise returned to living in the rearview mirror.

So, improvement stayed an event instead of becoming a capability. The shaping enterprise is the one that breaks that pattern. It builds a mechanism where improvements persist because they become encoded into how decisions are made and executed. The organization does not need to relearn the same lesson every quarter. It can store learning as decision logic. It can deploy that logic at the edge. It can detect drift early and intervene fast. That is not a dashboard project. That is not a consultant program. That is a new operating model. It is also why causal is not just an analytics preference. It is the only approach that can make that operating model real, because causality is the language of intervention. Correlation is the language of description.

The Shaping Choice

A CEO does not need another sermon about speed. A CEO needs a way to build speed that does not create fragility. The forward-facing alternative is not to throw away benchmarking. It is to demote it. Benchmarking becomes a hypothesis generator, not a strategy engine. Correlation becomes a way to ask better questions, not a way to copy answers. Then the CEO does something different. The CEO stops asking “what do leaders do,” and starts asking “what must be true in our operating system for us to produce the outcomes we want.” That question forces a different kind of work. It forces you to identify the causal structure of your performance. It forces you to model the constraints that actually bind you. It forces you to surface where your permission architecture is blocking action. It forces you to decide where to place autonomy and where to place gates, based on risk and leverage, not on tradition. It also forces a new honesty about where the enterprise is losing time. Time is the currency you are spending without budgeting. Every approval gate is a time withdrawal. Every escalation is a time withdrawal. Every meeting that exists because the system cannot settle truth is a time withdrawal. That time becomes cost. That cost becomes margin leakage. That margin leakage becomes strategic weakness. Eventually it becomes acquisition vulnerability. So the CEO level work is to collapse decision latency. Not by demanding urgency. By redesigning the operating system so that decisions can be made earlier and held stable. That is the staircase problem. It is not a culture problem first. Culture follows mechanism more than most leaders want to admit.

When permission is clear, conflict drops. When decision rights are encoded, escalation drops. When action is safe at the edge, meetings drop. When feedback loops are fast, trust rises because learning is visible and objective. People stop fighting about whose interpretation is correct because the system can test reality. That is how causality becomes a forward-facing capability. It makes the enterprise capable of steering, not merely observing.

The Unveiling

Now we can name the quiet truth that most CEOs can feel, even if they have never said it out loud. The dominant consulting narrative about productivity and standout firms is often brilliant at correlation and weak at control. It describes what winners look like. It does not redesign what followers are. It points to bold moves. It does not remove the architectural constraints that make bold moves impossible to execute without chaos. It celebrates standouts. It does not build the decision runtime that creates standouts. The specific report from McKinsey Global Institute, is a strong example of this genre. It is valuable as observational work. It is also rearview mirror work. It can tell you that a small number of firms account for an outsized share of productivity growth. It can categorize the kinds of moves those firms made. It can argue that productivity is not only about incremental efficiency. What it cannot do is give a CEO a forward-facing mechanism for shaping outcomes, because it leaves human inference and decision gates in place. It leaves the enterprise operating model untouched. It leaves the learning speed limit intact. That is why, in the correlation economy, benchmarking becomes surrender. You end up copying the visible correlates of winners while remaining structurally late. You become excellent at explaining why you are behind, which is not the same thing as changing what happens next. The shaping enterprise makes a different choice. It uses correlation to ask better questions, then uses causality to build the control system that answers them. It collapses decision latency. It redesigns permission. It turns improvement into a capability. It moves forward while others are still describing yesterday. That is the difference between imitation and shaping. And that is why the five-year nameplate test is merciless.

Because in five years, the market will not ask whether your leadership team was smart. The market will not ask whether your consultants were persuasive. The market will not ask whether your dashboards were beautiful. The market will ask who learned faster, acted earlier, and stabilized advantage into an operating reality. Then it will put a different name on your assets.

References

This article is grounded in The COO Council’s doctrine on decision latency, the learning speed limit, and the permission staircase, including the arguments developed in “The Hidden Speed Limit Inside Every Enterprise,” “The Decision Clock,” “The Architecture of Permission,” and the causal productivity maps you provided, which frame productivity as an accumulated advantage system that can be compounded or penalized depending on intervention speed and permission design. It is also informed by foundational work on bounded rationality and firm behavior, including Herbert Simon’s “A Behavioral Model of Rational Choice,” which explains why human decision making under constraint defaults to satisficing rather than optimization. JSTOR It draws on Cyert and March’s behavioral theory tradition on routines and negotiated reality inside firms, which helps explain why organizations revert to default decision patterns when architecture does not change. Tuck School of Business The forward facing argument for causality is aligned with Pearl and Mackenzie’s framing of structural causal models and counterfactual reasoning as the basis for intervention rather than description. UCLA Cognitive Systems Laboratory The control system lens is reinforced by Ashby’s Law of Requisite Variety, which clarifies why a regulator must match the environment’s variety and speed, or else it will purchase stability through delay. Ashby Digital Archive The learning loop emphasis is supported by Deming’s articulation of PDSA as a mechanism for continual learning and improvement, which is fundamentally about testing and updating theory through action, not narrating outcomes after the fact. The W. Edwards Deming Institute+1 Finally, the correlation based “standout firm” framing used as the illustrative rearview mirror example comes from McKinsey Global Institute’s “The power of one: How standout firms grow national productivity,” along with its supporting materials, which emphasize concentration patterns and observed moves but, by design, do not re-engineer enterprise decision runtime and permission architectures. McKinsey & Company+3McKinsey & Company+3McKinsey & Company+3