KNOWLEDGE · TOPIC PAGE · STATE OF RESEARCH
Organisational Transformation
Concept, evidence and methods in the current state of research 2026, compiled by
Bernhard Nitz · transformind GmbH · Kilchberg
"We're transforming." Few phrases come up more often in board and executive meetings. Fewer still mean the same thing twice. Sometimes it describes a new ERP rollout. Sometimes the repositioning of a business model. Sometimes it is simply the hope that patterns which have been stuck for years will finally give way.
Research draws the line more sharply than everyday usage suggests. It speaks of organisational transformation only when something changes at the core of a company: how it creates value, how it makes sense of its market, and how it understands its own role within it. Many initiatives that are called transformations internally are substantial change programmes. They can be demanding, expensive and strategically important. Yet not every far-reaching change reaches down to the level of identity and value creation.
Anyone deciding on transformation at board level is therefore working in a field where terms blur quickly. That is where this page begins. It condenses the state of research on organisational transformation, places the major schools of thought in relation to one another, and separates what is well established from what is merely often repeated. It shows why the familiar claim that seventy percent of all transformations fail rests on no traceable original study, what research and practice actually know about how deep change succeeds, and why so many initiatives stall at a point most models barely address.
CONCEPT
Transformation changes what an organisation is
Organisational transformation belongs to that class of terms which appear routinely in strategy papers, executive workshops and board meetings. What stands out is how rarely anyone agrees on what the term actually covers. For some it means a comprehensive restructuring. For others, the digitalisation of core processes. Others again use it for a shift in culture and leadership.
Research draws the boundaries more precisely. Across very different theoretical traditions, a striking consensus emerges: transformation begins where the core of an organisation changes. Three characteristics help place a given initiative.
Depth of change
Transformation alters more than processes, structures or systems. It alters the patterns through which an organisation reads its environment, reaches decisions and creates value. What was taken for granted yesterday stops being self-evident.
Scope of change
A transformation affects the organisation as a whole, or at least those parts of it that shape its future development. Individual functions can go through profound change without the organisation transforming.
Target state
Conventional change initiatives usually work towards a largely known outcome. In a transformation, the target picture tends to take shape along the way. The route itself produces the insights that sharpen the destination.
Which gives the working definition used throughout this page:
Which gives the working definition used throughout this page:
Working definition of organisational transformation
Organisational transformation is a far-reaching process of change that affects the whole organisation or substantial parts of it, alters its value creation logic, its interpretive frames or its identity, and whose target picture takes concrete shape only as the process unfolds.
The definition deliberately avoids normative assumptions. It implies neither success nor controllability. Both are empirical questions. Anyone who has carried responsibility for a transformation knows from experience that deep change can be neither fully planned nor fully controlled. The decisive question is therefore less whether transformation can be steered, and more which forms of leadership, decision-making and organisational learning hold up under which conditions.
The orders of change give the sharpest criterion
Board and executive discussions tend to focus on the size of an initiative. How large is the budget? How many people are affected? How long will it run? For the question of whether an organisation is undergoing change or transformation, these criteria are of limited use.
What matters is the level at which the change takes place, and precisely not the size of the initiative.
The theoretical groundwork was laid in the 1970s and 1980s (Golembiewski, Billingsley and Yeager 1976; Bartunek and Moch 1987). The central distinction still holds: change can happen within existing patterns of thought, or it can alter those patterns themselves.
First-order change improves an organisation within its established logic. Processes become more efficient, systems more modern, workflows leaner. The underlying assumptions about how the company creates value and why it succeeds remain untouched.
Second-order change cuts deeper. It alters an organisation's interpretive frames. Long-held certainties come under question, established formulas for success lose their validity, new forms of value creation emerge. The organisation changes more than what it does. It gradually changes how it reads its environment and what role it takes within it.
Some authors describe a third order beyond this: an organisation's capacity to reflect deliberately on its own patterns of thought and interpretation and to adjust them when needed. Here the focus shifts from the single transformation to the capability for continuous renewal.
For practice this yields a simple but consequential distinction. Change management largely shapes first-order change. Organisational transformation operates at the second order. Building long-term adaptability aims at the third.
This difference is more than academic. It determines what kind of leadership, planning and steering makes sense.
Consider an industrial company introducing a new ERP system. The project demands heavy investment, ties up leadership capacity for years and changes the working day of many employees. The company's underlying logic nevertheless stays intact: it still develops, produces and sells the same offerings to the same customers. The undertaking is demanding. It remains first-order change.
The picture is different when a machinery manufacturer moves its business model from selling equipment outright to guaranteeing availability. Sales now sells performance across a lifetime of use rather than machines. Service, risk management, contracting and the P&L all follow a new logic. The question is no longer only how the company works, but increasingly what it is to its customers. That is where transformation begins.
For boards, one practical consequence follows. Before choosing a method, establish which order of change you are actually in. Many initiatives stall because second-order change is being run with first-order instruments. The activities are managed professionally, while the real change takes place outside the chosen steering model.
The identity criterion draws the finest line
Anyone discussing transformation at board level arrives sooner or later at the same question: when does a significant change actually become a transformation?
The answer lies less in the size of an initiative than in its effect on the identity of the organisation.
The question became especially sharp in research on digital transformation, where a practical boundary problem kept recurring. Companies invested heavily in new technology, automated processes and digitalised customer contact. Much of it was strategically important. Yet not every one of these initiatives changed the organisation in any real sense.
Research therefore proposes a straightforward criterion (Wessel et al. 2021): as long as technology supports or improves an existing value creation logic, this is technology-enabled change. Transformation applies only where technology helps bring a new organisational identity into being.
The distinction looks fine at first glance. In practice it is often decisive.
A bank that makes its existing services available digitally remains, to its customers, the same bank. An industrial firm using sensors and analytics to schedule maintenance precisely remains fundamentally the same supplier. The way of working changes. The identity stays largely stable.
It is different when a company begins to be something fundamentally other to its stakeholders than it was before. When a manufacturer becomes an operator. When a product company becomes a service company. When the real contribution lies in the continuous provision of a service rather than in the sale of an object. In such situations more than the business model changes. The organisation develops a new understanding of its own role.
In family-owned companies this difference is often especially visible. Many changes touch processes, systems or structures. The real tension arises where the question surfaces of whether the company will still be the same company it has been. At that point transformation becomes a question of identity.
For steering, the distinction has considerable consequences. Initiatives with a stable identity can often be run through clear target pictures, defined programmes and conventional delivery methods. Where the identity of the organisation itself shifts, that logic usually stops being sufficient. The target picture develops in stages. New meanings emerge as the change unfolds. Leadership then consists less in delivering a finished state than in providing orientation during a period when core assumptions are being renegotiated.
A second distinction becomes important later on this page. Some transformations run as deliberately initiated programmes with a recognisable beginning and end. Others emerge from a multitude of small adjustments that condense over years into a new reality. Whether an organisation is going through episodic or continuous change (Weick and Quinn 1999) shapes its steering logic at least as strongly as the question of identity; the section on typology picks up that thread.
Two neighbouring terms cause recurring confusion
In conversations with boards and executive teams, organisational transformation, organisation development and restructuring are frequently named in the same breath. The terms do overlap in part. They describe different things, and the distinction is not incidental to how you proceed.
Organisation development denotes a tradition of change work and therefore fixes no particular type of change. Its starting point is the assumption that organisations become more resilient when the people affected take part in analysing and shaping the change. Many practices that are now taken for granted come from this tradition, from workshop formats through feedback processes to forms of team development.
An organisation can transform itself using the instruments of organisation development. It can equally use those same instruments to improve processes, strengthen collaboration or develop leadership practice. Organisation development therefore describes an approach, not the depth of the change.
Much the same applies to restructuring. Restructurings alter structures, accountabilities, cost bases or legal entities. They can have considerable consequences for a company; empirical research on reorganisation shows correspondingly heterogeneous performance effects that depend heavily on starting position and execution (Girod and Whittington 2017). In many executive teams they first become visible as new org charts, adjusted reporting lines or the reorganisation of business units.
Even so, the same caution applies: not every restructuring is a transformation.
A company can redraw its entire organisational structure without changing how it creates value, how it reaches decisions or how it understands its role in the market. The visible structures change. The underlying logic largely persists.
Conversely, some of the deepest transformations are barely detectable in the org chart at first. They begin with altered decision premises, new forms of collaboration or a different reading of customers, markets and value creation. The structural adjustments often follow much later.
In family-owned companies this pattern shows up regularly. A new org chart does not yet create a new organisation. Nor does a culture programme on its own produce a transformation. What matters is whether the things that carry the company at its core actually change: its identity, its value creation logic and the assumptions on which its decisions rest.
The distinction is therefore more than terminological precision. It guards against mistaking scale for depth. A large restructuring programme can remain first-order change. A comparatively modest strategic repositioning can alter the foundations of the company and take on the character of a transformation.
For boards this yields a simple orienting question: are we primarily changing the structure of our organisation, the way we do change work, or the foundations of our business? Only the answer makes clear what is actually under discussion. The practical short form of these distinctions, including the question of which advisory approach fits when, can be found in the FAQ on the transformation page and in the FAQ hub.
EVIDENCE
What research shows about forming judgement
Anyone who reads seriously about organisational transformation meets one number with remarkable regularity: seventy percent of all transformations fail.
The claim turns up in keynotes, textbooks, consulting brochures and studies. Most executives know it. Some quote it from memory. Others have already made decisions on the strength of it. I used it myself for years and put it into my own articles, until at some point I grew sceptical. Because when you ask for the original empirical source, it goes remarkably quiet.
Hughes (2011) traced the origin of the number systematically. The result is sobering. The claim cannot be attributed to any sound primary study. Publications cite other publications, which in turn rest on further citations. Over the years, assumptions hardened into facts and estimates into settled knowledge.
What is striking is how durable the number has proved despite being questionable.
And there are understandable reasons for that.
For consultancies it raises the perceived urgency of professional help. For authors it makes an effective opening. And for executives it offers a measure of psychological relief. If the great majority of transformations fail, then one's own failure looks less like the consequence of specific decisions and more like statistical normality.
Which is exactly why the number deserves particular caution.
Research today knows no reliable base rate for how many organisational transformations actually fail. Definitions of success, the time horizons studied and the organisations examined vary far too widely.
That uncertainty does not make the field arbitrary. It simply moves attention to the better questions.
Against what is success measured? Adherence to budget and schedule? A change in behaviour and collaboration? An improved competitive position? Depending on the lens, the same initiative can be judged very differently.
The time horizon matters just as much. Most boards know projects that looked like disappointments after two years and were seen as landmark decisions a few years later. The reverse happens too. Early visible progress sometimes turns out, in hindsight, to have been an episode without lasting effect.
Judging a transformation seriously therefore takes more than a memorable rate. It takes clearly defined success criteria, an explicit time horizon, and a clean separation between assumption and observation.
This is precisely where research has moved forward in recent years. Studies of AI-driven transformation in particular are now producing work that can separate cause from effect, which allows a far more precise view of the conditions under which change succeeds. Those findings are covered in detail further down this page.
For boards and executive teams the consequence is simple: scepticism towards headline success and failure rates is a mark of good governance. Anyone deciding on transformation should ask less about universal probabilities and more about which conditions actually hold in their own company.
TYPOLOGY
Four engines drive organisational change
Transformation initiatives generate a great deal of debate about method. Should the company run a programme? Pilots? Agile teams? Culture initiatives?
The more important question is usually settled earlier: what forces are actually driving the change?
Misread that dynamic and you can run a methodologically immaculate initiative that works past the reality of the organisation. Research distinguishes four fundamental engines of change (Van de Ven and Poole 1995). In practice they rarely appear in pure form. The distinction still helps in understanding the logic of an initiative.
The first engine is purposeful change.
Here change arises from a deliberate decision. The board, the executive team or the owners define a goal and align the organisation towards it. Most strategy, restructuring and transformation programmes follow this logic. It matches the picture of leadership that most management books quietly assume: someone recognises the need for change, describes the goal and organises the path towards it.
The second engine is development along a predetermined path.
Organisations change partly because their development follows recognisable patterns. A family-owned business makes different demands on leadership and control when it grows from fifty to five hundred people. What worked in one phase eventually reaches its limits. The change is less triggered than required.
The third engine is conflict.
Many transformations shift power, resources and influence. That creates tension between units, functions and levels of leadership. The decisive turns are then set as much through negotiation, coalitions and the balancing of interests as through analysis and project plans.
Watch board and executive meetings over a longer period and the dynamic becomes visible. Differing positions often rest on differing interests, accountabilities and risk perspectives, while the facts are largely shared. Transformation then emerges from how those tensions are handled.
The fourth engine is adaptation through variation.
Some changes are neither centrally planned nor politically negotiated. They emerge through many local experiments from which new patterns gradually form. Individual teams develop new practices. Units find different ways of working together. What proves itself spreads. What does not hold disappears again.
With technological change this logic is especially easy to observe. While central programmes are still being debated, individual employees have already changed how they work. Part of the real transformation then takes place outside the formal architecture.
For boards, the practical value of this typology lies in one simple insight: not every transformation follows the same logic.
Many established models assume change is primarily purposeful. Experience shows that transformations in mid-sized and family-owned companies usually contain several forces at once. Strategic decisions meet existing power structures. Formal programmes meet local experiments. Planning meets emergence.
The stronger these different dynamics, the less a purely programmatic approach will suffice.
Before choosing a method, then, one diagnostic question is worth the time: which forces are actually moving this change, and which of them does our current approach account for?
The quality of that answer often influences the odds of success more than the choice of any particular transformation model.
A programme with a beginning and an end, or a rebuild of running routines
Anyone who accompanies transformations sees the same starting position again and again. The discussion turns on goals, milestones and projects. Far more rarely does anyone ask the basic question of what kind of change this actually is.
Not every transformation follows the same logic in time.
Some changes run as clearly bounded undertakings. There is a recognisable trigger, a defined mandate and a point at which the programme closes. Introducing a new production system, integrating an acquisition or repositioning a business division often follow this pattern.
Other changes develop over years. They emerge from a multitude of small adjustments in daily leadership, in collaboration, or in contact with customers and markets. No single initiative is transformative on its own. Taken together, these adjustments change the organisation substantially.
Research therefore distinguishes between episodic and continuous change (Weick and Quinn 1999; on situated, continuous change Orlikowski 1996).
Episodic change follows the logic of deliberate intervention. The organisation recognises a widening gap between how it currently works and what its environment demands, and responds with a targeted change initiative.
Continuous change follows a different dynamic. Change arises as an ongoing part of organisational life. Decisions, experiments and adjustments condense over time into new routines, structures and ways of thinking.
For boards and executive teams the distinction has practical weight.
Treat continuous change like a conventional project and the programme usually ends long before the real change has landed in the organisation. Treat a clearly bounded undertaking purely as a cultural or evolutionary process and the necessary commitment in delivery tends to be missing.
In practice most transformations contain elements of both. A strategic repositioning is often set in motion by a programme. Whether it becomes durable is decided afterwards, once new routines, leadership practices and decision paths have become part of normal operations.
This is precisely where a perception gap opens up in many organisations. For the project organisation, the transformation ends with the final milestone. For the units affected, the real change often starts there.
Asking whether transformation should be understood as a programme or as continuous change therefore rarely leads anywhere. The more useful question is which part of the change follows which logic.
The clearer that distinction, the more realistic expectations, resource planning and measurement become.
THEORY
Five lines of theory carry the field, and none is universally valid
The theory of organisational transformation has an ambivalent reputation. Practitioners often find it detached, and part of that scepticism is earned, because the change literature passes on some of its own foundations with surprising imprecision. Knowing the five load-bearing lines of theory nevertheless yields three insights that bear directly on how you steer, along with a healthy distrust of the textbook versions in which they usually arrive.
It begins with Kurt Lewin, and with him the first example of imprecise transmission. The popular three-phase model of unfreezing, changing and refreezing is treated as Lewin's legacy, yet the reconstruction of his original texts by Cummings, Bridgman and Brown (2016) shows that he never developed this staged model as the formal foundation of a theory of change. The canonical form was shaped by textbook authors after his death and hardened over decades into the linear sequence that is now both cited and criticised. The widespread critique of a rigid, static Lewin therefore lands on the reception rather than the original. His actual core, the idea of behaviour as the product of a whole field of forces and the notion that change is better achieved by weakening the restraining forces than by adding pressure, sits far more comfortably with modern process-oriented thinking than the textbook version suggests. For this page that is more than history of science. It grounds the working rule of checking load-bearing claims against the primary text rather than against their popular summary.
The second line describes transformation as an exceptional state. Punctuated equilibrium theory (Tushman and Romanelli; tested empirically by Romanelli and Tushman 1994) sees organisational development as long periods of quiet, incremental adaptation interrupted by short, upheaval-like phases. In those phases the deep structure of the organisation, meaning strategy, structure, distribution of power and control systems, is reordered simultaneously. Empirical testing in the early computer industry supported the pattern: fundamental transformations occurred predominantly in short bursts, and many small individual changes precisely did not add up to a rebuild of the deep structure. This theory offers the strongest argument against the common hope that transformation can be reached as the sum of many small, low-risk steps.
The third line pushes back. Orlikowski (1996) documented, in a software implementation, how an organisation changed substantially over two years through the continuous improvisations of its users, without any upheaval programme ever ordering it; Brown and Eisenhardt (1997) showed that firms with semi-structured processes, meaning clear accountabilities alongside open design, innovated continuously where both tightly planned and structureless competitors failed. The apparent contradiction with punctuated equilibrium resolves as a question of context. In stable environments with a settled deep structure the episodic pattern dominates; in highly dynamic ones, the continuous pattern does. For practice, the fork from the typology returns: upheaval followed by consolidation, or the building of permanent adaptation routines. Wanting both at once, meaning permanent deep change as the normal state, is theoretically incoherent and runs empirically into the limits of absorptive capacity described in the section on capacity.
The fourth line tries to capture the capacity for change itself. Dynamic capabilities theory (Teece, Pisano and Shuen 1997) describes an organisation's ability to integrate, build and reconfigure its resources deliberately, and it is the most influential framework on the subject. It carries a long-recognised design flaw, however. If the capacity for change is recognised only by its success, then the claim that it leads to success turns from a testable hypothesis into a definition that confirms itself. This circularity is no academic worry; it describes exactly the mechanism by which organisations and their advisers attribute success after the fact to their own method. The way out is measurement before the outcome, against observable quantities such as the frequency of structured resource reallocation or the survival of management practices under strain. The full treatment of this problem, with both the supporting and the limiting evidence, is handled by the topic page on adaptive organisations; this page takes its findings as given.
The fifth line describes a tension that shapes many transformations as a target architecture. Research on ambidexterity (March 1991; O'Reilly and Tushman) examines how an organisation can exploit its existing business while exploring new ground. For small and mid-sized companies the finding of Lubatkin and colleagues (2006) is particularly relevant: there the integrating work of the top team carries the load, because separate exploration and exploitation units are rarely affordable in resource terms. Ambidexterity therefore describes where a transformation should arrive, not how it proceeds. The full treatment with study cards is on the topic page for ambidexterity.
Behind all five lines stands the methodological standard set by Andrew Pettigrew in his longitudinal studies (1985, 1990). Change can only be understood as the interplay of content, meaning what changes, process, meaning how it changes, and context, meaning under which internal and external conditions, examined across time rather than as a snapshot. Against that triad the weakness of many success-factor lists becomes visible: factors without reference to process and context are heuristics at best. And from it follows the rule with which the section on diagnosis opens: before the choice of method comes the analysis of context.
LEADERSHIP
The critical leadership work sits with the top team
Most accounts of transformation follow a simple dramaturgy. A leader recognises the need for change, develops a compelling vision and leads the organisation into a new future.
Reality is usually less cinematic.
In most companies the course of a transformation is decided by the quality of collaboration in the top leadership body, and rarely by any single leader. That is where priorities are set, resources moved, trade-offs surfaced and uncomfortable decisions taken. It is also where the consistency arises that an organisation needs in order to orient itself under uncertainty.
Sit in on board and executive meetings over a longer period and a recurring pattern emerges. Transformations rarely stall because the goal is unclear. More often they lose force because the leadership team holds different views about what is to be protected, what changed and what given up.
Research paints a similar picture (on the methodological critique of the long-dominant construct of transformational leadership, Van Knippenberg and Sitkin 2013).
The success of deep change depends less on individual leadership styles than on a team's ability to bring different perspectives together productively while remaining able to act. This matters most where established formulas for success have to be questioned. In those situations formal agreement is not enough. What counts is the ability to make differing assessments visible and still reach decisions that hold.
Transformation generates a particular kind of tension here. On one side the organisation expects orientation. On the other, the very assumptions on which that orientation rested have come loose. Leadership under those conditions means staying capable of decision even while central questions remain open, rather than conveying complete certainty.
For mid-sized and family-owned companies this adds a further difficulty. Strategic decisions touch not only commercial interests but history, identity and the company's sense of itself. Discussions about new business models, succession, internationalisation or digital value creation are therefore rarely purely factual. They tend to touch the question of what the company should become.
The real leadership work lies in neither dramatising nor suppressing these tensions. Leadership bodies that carry a transformation well make room for differing readings of the situation without drifting into indecision. They tolerate disagreement without losing the ability to decide. And they create commitment without projecting certainty where uncertainty remains.
From this angle the role of communication shifts too. Employees expect comprehensible decisions, recognisable priorities and visible alignment between words and actions; nobody is asking for the illusion of complete clarity. Where that alignment is missing, the organisation loses confidence in the change regardless of how good the programme is.
For steering a transformation, one plain observation follows: the capability of the top team is part of the change itself.
Which is why diagnosing a transformation often does not begin with employees or in the functions. It begins with the question of whether the leadership body has the clarity, coherence and capacity for disagreement needed to carry an organisation through a period of fundamental change.
Why the load shifts onto the leadership team
As long as a company operates within an established business model, much of the work of leadership can be organised along functions and accountabilities. Sales owns the market, operations owns delivery, finance owns the allocation of resources.
Transformation disturbs that order.
New business models, digital value creation, changed customer expectations or technological disruption almost always cut across existing boundaries of responsibility. Decisions in one area suddenly have significant effects in another. The consequence: the decisive questions can no longer be answered sensibly inside individual functions.
The load of leadership shifts onto the top team.
In practice this happens earlier than those involved tend to realise. The discussion appears to be about investment, priorities or organisational design. Behind it sit more fundamental questions. Which parts of the existing business model should be protected? Where is deliberate cannibalisation acceptable? Which risks are tolerable and which are not? And which tensions have to be held rather than resolved prematurely?
No single executive holds all the relevant perspectives.
Transformation demands the ability to bring different views together without flattening them. Sales sees the market. Operations sees feasibility. Finance sees what the company can carry. Owners see long-term continuity. In family-owned companies, historical and identity-related perspectives are often added that do not translate into a business case.
The capability of a leadership team therefore shows less in how quickly it agrees than in the quality of its shared judgement.
Observe leadership bodies over time and a recurring difference appears. Weaker teams reach for consensus quickly. Stronger teams make room for differing readings before they decide. The reason is rarely a greater appetite for conflict. These teams have understood that complex change can seldom be understood from a single vantage point.
Research supports the observation (Hambrick on behavioural integration in top teams; Edmondson 1999 and Frazier et al. 2017 on psychological safety). What seems to matter is less the harmony of a leadership team than its ability to process information jointly, share responsibility and stay capable of action despite differing views.
One further factor is routinely underestimated: the safety to voice disagreement.
In many leadership bodies, critical assessments exist long before they are spoken aloud. Strategic risks are known. Doubts about assumptions are held. Reservations about initiatives are present. The difference lies in whether the system makes those perceptions visible; they are almost always there.
Transformation raises the value of that capability. The greater the uncertainty, the more expensive unspoken objections become.
At the same time there is little to be said for treating conflict as an end in itself. The popular advice that good leadership teams mainly need to argue more falls short (De Dreu and Weingart 2003). Leadership bodies improve when different perspectives are raised early, examined seriously and translated into decisions that hold; louder conflict alone does not achieve that.
For boards and executive teams this yields a practical diagnostic question: how well does the leadership team surface differing views before decisions are made?
The answer often says more about an organisation's capacity for transformation than any org chart, roadmap or transformation programme. How a leadership team works on this is covered on the page for team development.
Decision quality is a steering variable in its own right
Transformations rarely follow a finished plan. They develop through a sequence of decisions that have to be taken under uncertainty.
New information appears. Assumptions turn out to be incomplete. Markets move differently than expected. Technological development speeds up or slows down. The quality of a transformation therefore depends not only on its strategy or its programme but on the quality of the decisions taken along the way.
That perspective changes the view of leadership.
In many companies decisions are still judged primarily by their outcome. If an initiative develops well, the decision looks right. If success fails to appear, it is treated in hindsight as a mistake.
For leadership bodies this logic is a problem. Good decisions can produce poor outcomes. Poor decisions can look successful for a while. Confuse the two and you tend to draw the wrong lessons from the past.
Decision quality and decision outcome are not the same thing.
It is therefore worth examining the quality of a decision independently of how it later turns out (Spetzler, Winter and Meyer 2016). Were the relevant alternatives examined? Were the underlying assumptions made explicit? Was the available information sufficient? Were risks and trade-offs surfaced? And was there clarity about the measure against which the decision would later be judged?
Watch leadership bodies and a recurring difference appears between robust and fragile decisions. The better decisions rarely emerge because more information was available. They often emerge because different perspectives were examined more systematically.
In transformations this matters especially. New business models, digital technologies or changed market conditions often lie outside prior experience. The uncertainty sits not only in the answers but already in the questions being asked.
Which is why the quality of shared judgement gains weight.
An effective leadership body examines more than its preferred option. It engages equally with the plausible alternatives. It questions the assumptions behind a decision. And it creates room for reasoned counter-positions before consensus sets.
In practice it often turns out that critical signals were present early. They were heard but not examined closely enough. Or they went unspoken because the decision was treated as settled. The cost of such omissions usually becomes visible only once a change of direction has grown considerably more expensive.
For open-ended transformation work, a further consequence follows: not every decision has to be final.
The more uncertain the environment, the more sense a steering logic makes that works with testable assumptions. Load-bearing hypotheses are named. Leading indicators are defined. Observation points are fixed. And it is settled in advance under which conditions a chosen path will be confirmed, adjusted or abandoned.
For boards and executive teams this is among the most demanding leadership tasks there is: organising the capacity to decide under uncertainty, rather than manufacturing the illusion of safety. Transformation does not become predictable through this. It does become testable.
And that is precisely the difference between hope and steering. The architecture behind it is developed on the topic pages for decision quality and judgement in the AI age; the institutionalised form of external challenge is described on the sparring page.
COMMUNICATION
Communication is settled in the daily conduct of leadership
Communication is among the most discussed topics in any transformation. It is also among the most misunderstood.
In many organisations communication is treated as a task that begins once the substantive decisions have been made. Presentations get built, information sessions scheduled, messages drafted. The real work of leadership is located somewhere else.
Practice argues for a different view.
Transformations rarely fail for want of information. In most companies employees know very early that change is coming. Uncertainty arises less from missing news than from open questions. What does this mean concretely? Which of our previous assumptions still hold? What is expected of us? And how would the organisation itself know whether it is on the right track?
Communication serves a different function here than many programmes assume. It helps people make sense of a new situation, which reaches well beyond the transfer of information.
Accompany leadership bodies through larger changes and the same dynamic appears. Discussion inside the organisation rarely centres on the official messages. It centres on the gap between those messages and the decisions people can observe.
Employees hear what is said. They orient themselves more strongly by what gets priority, which decisions are taken and which behaviour is actually rewarded.
Which is why the credibility of a transformation is only partly settled in town halls, strategy decks or value statements. It is settled in the daily conduct of leadership.
Research confirms the observation (Oreg, Vakola and Armenakis 2011). Perceived fairness, comprehensible decision paths and genuine opportunities to participate shape the response to change more strongly than the sheer frequency of communication. People accept difficult decisions more readily when they experience the process behind them as intelligible.
From this follows an important distinction. Information creates clarity about facts. Orientation arises only where connections become intelligible.
In transformations that orientation cannot be produced completely. Many questions are still open at the outset. Business models develop. Priorities shift. New findings change the original plan.
The task of leadership is therefore to make uncertainty as transparent as possible while remaining able to act, rather than simulating certainty.
In practice this often means saying things that classic change programmes tend to leave unsaid. That not all the answers are in. That target pictures are under review. That assumptions may be adjusted. And that certain questions are being deliberately held open.
Paradoxically, this kind of openness often strengthens trust, because it acknowledges the reality employees are already perceiving.
One further observation deserves attention. Participation works only where it allows real influence. Where decisions are communicated first and then discussed under the heading of participation, the intended effect regularly reverses. The organisation reacts less to the absence of a say than to the impression that a say is being simulated.
For boards and executive teams this yields a simple guiding question: which uncertainties are allowed to be spoken about openly in our organisation?
The answer often says more about the communicative capability of a transformation than the number of communication measures or the polish of their presentation.
In the end, communication is how an organisation produces meaning while the order it has known is changing. Treat it as a supporting measure and you have underestimated it.
ACCEPTANCE
Acceptance comes from how the change is designed
Transformations are usually described from the perspective of those who initiate them. Strategies get developed, programmes set up, measures defined. Far more rarely does attention turn to the people who have to work with the consequences of those decisions.
Yet a substantial part of the effect is decided precisely where those affected are brought into the design.
Research on responses to change has produced a remarkably consistent picture over recent decades (Oreg, Vakola and Armenakis 2011). Resistance rarely arises because people are fundamentally against change. It arises more often where change is experienced as unintelligible, unfair or beyond influence.
The distinction has practical weight for leadership bodies.
In board and executive meetings, acceptance is often discussed as a property of the organisation. The question becomes whether employees are ready for the change. Research suggests inverting the perspective. The more useful question is often whether the change has been designed so that readiness can arise at all (on measuring readiness for change, Armenakis, Harris and Mossholder 1993; Holt et al. 2007).
People respond to the concrete consequences a transformation has for their working day; as an abstract concept it reaches almost no one.
Does my role change? Do I lose influence? Are new demands coming? Will the experience I have built still be valued? Will relationships formed over years survive? Questions like these shape the perception of change far more strongly than strategic arguments or market developments.
Accompany managers and employees through a transformation and the same divergence appears. What looks at the level of strategy like a logical next step is often experienced at the level of daily work as loss, uncertainty or added burden.
None of these perceptions is irrational as such.
Acceptance therefore arises less through persuasion than where people can follow why a change is necessary, what contribution they can make and how decisions come about.
Procedural fairness carries particular weight here.
Organisations often accept difficult decisions more readily than unclear or inconsistent ones. Even severe changes can be carried when the criteria appear intelligible and accountability stays visible. Conversely, even minor interventions cause irritation when the reasoning and the decision path remain in the dark.
A second finding deserves attention. Participation works only where there is genuine room to shape the outcome.
Many change programmes speak of involvement although the substantive decisions have already been taken. In such situations additional acceptance rarely appears. Often the opposite does. The organisation reacts less to the decision itself than to the impression that participation is being staged.
For leadership bodies this holds an uncomfortable but important insight: honest leadership creates more trust than performed openness.
Research also shows that readiness for change is not available without limit. Organisations have a finite capacity to absorb new initiatives, priorities and programmes. Exceed that limit for long enough and fatigue, withdrawal and cynicism towards further change set in.
Most managers recognise the phrasing from their own corridors. "Let's wait and see." Or: "This too shall pass." Such reactions are rarely a sign of poor motivation. Often they indicate that the organisation has already used up its capacity to process.
For steering a transformation, one central consequence follows.
Acceptance cannot be mandated; it arises as the result of fairness, intelligibility, participation and a pace of change the organisation can actually absorb.
Treat acceptance purely as a communication task and you have missed its actual causes. Understand it as a consequence of how decisions are made and change is designed, and you gain a far more precise lever for leading transformations.
In that sense acceptance is an indicator of how the organisation is experiencing the change while it happens, and therefore much more than a downstream success factor.
CAPACITY
An organisation's capacity to absorb change is finite
Most leadership bodies discuss resources intensively. Capital, capacity, talent, market position.
Far more rarely do they discuss a resource that matters at least as much for transformation: the organisation's ability to take in and process change. That ability is finite.
The observation may seem obvious. In practice it is routinely underestimated. While individual transformation initiatives are planned carefully, a portfolio of parallel initiatives tends to accumulate over the years. Strategic programmes, reorganisations, system implementations, culture projects, digitalisation initiatives and regulatory adjustments all run at once.
Each initiative looks reasonable in isolation. Together they not infrequently overwhelm the same organisation.
Work with managers and employees over time and certain patterns recur. Openness to new initiatives declines. Projects are formally supported but only superficially implemented in daily work. Priorities change faster than they can be processed. And in place of active engagement a quiet form of waiting sets in. Research describes these patterns as excessive change and change fatigue (Stensaker et al. 2002; Bernerth, Walker and Harris 2011).
The cause often lies less in missing motivation than in overstretched absorptive capacity.
Organisations process change through more than structures and processes. They process it through people. New roles have to be understood. New habits have to form. New ways of working together have to be learned. Every change consumes attention, energy and time. Those resources cannot be parallelised at will.
In mid-sized and family-owned companies this shows up particularly clearly. The same manager who is carrying a transformation project usually also owns customer relationships, operational decisions and the leadership of their own unit. Change happens inside the day job and cannot be handled alongside it.
For leadership bodies this yields a consequence that is often more uncomfortable than deciding on a new initiative: deciding against another one. Transformation demands not only addition but subtraction.
Every new priority raises the question of which existing priority loses weight. Every additional initiative claims attention that is then unavailable elsewhere. Organisations that do not actively steer this displacement tend to experience change as compression rather than development.
A further error lies in the notion of permanent transformation. Seen from a strategy paper, continuous change often looks desirable. Seen from the perspective of organisational resilience, a more differentiated picture emerges.
Change needs phases of consolidation.
New routines have to stabilise. Managers have to accumulate experience. Employees have to gain confidence about what is actually expected. Without such consolidation phases, the likelihood grows that each new initiative meets an organisation whose attention is already committed elsewhere. The underlying diagnostic pattern, treating perpetual change as a virtue, is developed further on the topic page for adaptive organisations.
For boards and executive teams this produces a central steering question: how much change is the organisation actually carrying right now?
The question looks unspectacular at first. In practice it often decides whether a transformation can take effect or simply becomes part of an increasingly crowded landscape of initiatives.
The capacity for transformation is therefore not measured by willingness to change alone. It is measured equally by the discipline to prioritise change, sequence it and see it through.
CULTUR AND POWER
Culture, identity and power are harder than their label
Many transformation programmes describe culture, identity and power as soft factors.
Anyone who has watched change in organisations over time tends to reach the opposite conclusion.
Strategies can be decided. Structures can be adjusted. Processes can be redesigned. Culture, identity and power respond more slowly. Which is exactly why, over the long run, they often do more to determine whether a transformation holds or peters out.
Culture rarely shows itself where it is discussed.
It shows in the decisions that get taken. In the trade-offs that are accepted or avoided. In the behaviour that earns recognition. And in the topics that are deferred again and again.
Which is why culture programmes on their own often change little. New values on a slide do not yet make a new culture. What counts is whether leadership behaviour, decision paths, incentives and ways of working actually change. The field study by Canato, Ravasi and Phillips (2013) on the introduction of Six Sigma at 3M shows the mechanism in detail: the shift in culture came about through the sustained practice of new routines, whose sense the organisation worked out only in the doing.
Follow board and executive meetings and the connection becomes visible. Organisations change their culture through repeated decisions that make new behaviour credible; communication alone rarely achieves it.
A similar dynamic applies to organisational identity.
Every company develops over time an understanding of who it is, why it succeeds and what it contributes. This self-understanding is rarely discussed explicitly. Which is precisely why it tends to act so strongly.
Transformations touch this level earlier than project plans reveal.
When an industrial company moves from selling products to service or platform models, more than the value creation changes. What changes is the answer to the question of what the company actually is to its customers.
Many conflicts in transformations therefore arise over the meaning of a change and only rarely at its surface. The stated discussion concerns a new business model or a new technology. The real question is often whether the company is prepared to question how it has understood itself.
In family-owned companies this level plays a particular role. History, tradition and identity are more than an emotional reference point there. They often form a substantial part of the organisation's stability. Successful transformations therefore build on the existing identity and develop it further, rather than working against it (Gioia and Chittipeddi 1991).
The third factor stays largely invisible in most transformation models: power.
Change shifts influence, resources and room to act. Where new priorities emerge, existing ones lose weight. Where new accountabilities emerge, existing power relations shift.
This dynamic is neither unusual nor problematic. It becomes problematic only when it is ignored.
In practice much resistance is prematurely interpreted as a lack of willingness to change. Often it is a comprehensible response to real shifts in influence, responsibility or standing.
Transformation is therefore never solely a technical process. It is always a social and political one as well (foundationally, Pettigrew 1985).
For boards and executive teams this yields a central diagnostic question: which cultural certainties, identity assumptions and power structures does this transformation touch?
The answer often explains more about the dynamics of an initiative than the quality of its project plan. That question about power is also the positioning core of the transformation page, which treats culture change for what it is.
What gets labelled a soft factor turns out in practice to be the load-bearing structure beneath the visible organisation.
PERCEPTION
Leadership and organisation often experience the same transformation differently
Many transformations run over months or years. In that time a quiet divergence usually opens between how leadership perceives the change and how the organisation does.
Both sides are looking at the same thing. They still arrive, not infrequently, at different assessments of progress, effect and success.
For leadership bodies the observation matters particularly.
Boards and executive teams receive their information largely through programmes, metrics, steering committees, reports and milestones. That produces a picture of the transformation shaped strongly by planning and delivery.
Employees experience the same transformation from a different position.
They judge change by their working day. Have decision paths changed? Does collaboration work differently than before? Are new priorities actually being lived? Has the promised change become visible in daily work?
Both views are legitimate. They are measuring different things.
Which is exactly why transformations keep producing situations that surprise leadership bodies. Programmes run to plan from the steering perspective while the organisation perceives little progress. Conversely, employees report profound change although key milestones have formally not been reached.
The reason lies in the nature of organisational transformation.
Deep change often shows up in people's experience earlier than in a programme's metrics. At the same time, programmes can generate considerable activity without the underlying patterns of the organisation having shifted at all.
Accompany transformation work over time and a recurring pattern emerges. The greatest risks rarely sit where leadership and organisation read the situation the same way. They sit where the two perceptions drift apart without that gap becoming visible.
This gets critical in phases where uncertainty, workload or strategic tension increase. Leadership then sometimes reads missing progress as a delivery problem. The organisation experiences the same situation as the consequence of unclear priorities or contradictory signals.
For boards and executive teams this has a simple but consequential implication: progress should never be judged on programme data alone.
Metrics, milestones and project reports carry important information. They show what has been delivered. They show only in part how the change is actually landing in the organisation.
Nor is it enough to rely on mood readings or individual observations. Perceptions show how change is experienced but say nothing automatic about its strategic effect.
The most robust view emerges where both perspectives are brought together.
Programmes supply information about delivery. The organisation supplies information about effect. Only the interplay of the two produces a picture solid enough to steer deep change.
For leadership bodies a useful diagnostic question follows: where do our assessment of the transformation and the organisation's perception diverge?
Often it is precisely in that gap that the most telling indication of a transformation's actual state can be found.
METHODS
Comparing methods ends in a logic of fit
Anyone working on organisational transformation quickly runs into a profusion of methods, frameworks and schools. The range runs from classic change models through organisation development and systemic consulting to agile transformation approaches.
The debate then tends to follow a familiar pattern. Which method works best? Which approach has proved itself? Which model should we use?
Research suggests a different view.
The decisive question is rarely which method is better in general. What matters is under which conditions a given approach holds and where its limits lie.
That perspective also matches the reality of most transformations.
Boards and executive teams understandably look for orientation. A model creates structure. A defined approach conveys safety. At the same time, the starting positions of organisations differ considerably. A family business in succession faces different problems than an industrial company in digitalisation or a service firm reordering its business model.
Against that background, the idea of a universally correct method looks increasingly fragile.
Some approaches assume the starting position can be analysed well enough to plan the route of change. Others treat organisations as social systems whose development can only be steered from outside to a limited degree; research has worked out the distinction between these diagnostic and dialogic stances precisely (Bushe and Marshak 2009). Others again rely on experiments, learning and stepwise adjustment. How transformind practises the systemic, complementary approach in its own work is described on the approach page.
Each of these perspectives holds valuable insight.
How far each carries depends on the kind of change at hand.
Where the target picture and the route are largely known, the greatest value usually comes from clarity, reliability and disciplined delivery. Conventional project and programme management can be highly effective under those conditions.
The deeper a change cuts into identity, culture, value creation logic or power structures, the more the limits of purely plan-driven approaches come to the fore. The relevant questions then often cannot be answered fully in advance. They become visible as the change unfolds.
Many leadership bodies know the experience. The decisive insights often emerge during delivery rather than during planning. Hypotheses are confirmed or lose plausibility. New tensions surface. Connections that were previously overlooked become visible.
Transformation thereby takes on a reflexive character. The organisation changes more than its actions. Step by step it also changes its understanding of the situation.
Which is why arguments about method tend to run out of road in practice. The more relevant capability usually lies in combining perspectives sensibly, and far less often in applying a single model consistently.
A robust transformation typically holds several logics at once. It uses analysis where analysis is possible. It relies on participation where acceptance is decisive. It works with experiments where uncertainty stays high. And it produces clear decisions where orientation is needed.
For boards and executive teams a sober consequence follows.
The quality of a transformation is rarely measured by which model it follows. It is measured by whether the chosen approach fits the actual dynamics of the organisation.
Methods offer orientation. Responsibility for the fit stays with leadership.
Which is why comparing methods ends in a diagnostic question rather than a ranking of approaches: what kind of change is this, and what form of leadership matches its logic? How the Ambiflow diagnostic and steering framework relates to Kotter's well-known eight-step model is answered in the FAQ on the transformation page.
DIAGNOSIS
Diagnosis comes before the choice of method
Many transformations begin with the search for a solution.
A new operating model is presented. A methodology is selected. A programme is set up. Often this happens before there is sufficient clarity about what kind of change is actually at hand.
This is where a good share of what later appears as delivery problems originates.
In leadership bodies the same pattern shows up regularly. Discussion moves quickly to the level of possible measures. Considerably less time goes into understanding the starting position precisely. Yet that groundwork decides whether a method holds or works past the reality of the organisation.
Transformation therefore begins with a diagnosis, and the roadmap follows afterwards.
Four questions sit at the centre:
-
The first question concerns the depth of the change.
Is this an adjustment within existing formulas for success, or a change to the assumptions on which the company has been built? Put differently: is this first-order change, or transformation in the proper sense? -
The second question concerns the forces moving the change.
Does the change arise primarily from a strategic decision? Is it driven by technological development, market shifts or new regulatory requirements? Or do competing interests, shifts in power and local experiments shape the dynamic more strongly than the original plan? -
The third question concerns readiness for change.
This is not about agreement in the narrow sense. What matters is whether the organisation currently has the preconditions to absorb deep change at all; validated instruments exist for measuring this (Holt et al. 2007; Bouckenooghe et al. 2009). Readiness is also rarely distributed evenly. What appears obvious in a leadership group can be perceived very differently in operational units. -
The fourth question concerns the existing change load.
Every organisation has finite attention. Running programmes, restructurings, system implementations and strategic initiatives all draw on the same people. Plan a transformation without making that load visible and you will usually overestimate the organisation's real capacity to deliver.
These four questions may look unspectacular. In practice they explain a considerable share of the difference between robust and fragile transformation architectures.
A good diagnosis answers more than the question of what to work on. It also creates clarity about how progress will become recognisable in the first place.
And here sits a frequently underestimated difficulty.
Many organisations measure transformation through outcomes that only become visible at the end. Growth, productivity, profitability or market share are central figures. For steering a transformation as it runs, they are of limited use. They show effect, often only once the actual causes lie well in the past.
Which is why indicators that become visible earlier gain importance.
It is observable, for instance, how consistently resources are reprioritised. It is observable whether central leadership and decision routines survive under the pressure of change. And it is observable whether investment goes not only into the technology but into the people and processes meant to carry it: into data, process design and capability building. Such indicators say nothing about later success, and that is precisely their strength. They show early and independently of the outcome whether the organisation is doing what makes change likely, and they resist attribution after the fact.
In transformind's practice this logic is built into the Ambiflow diagnostic and steering framework and its Pulse Check, which assesses six conditional dimensions of organisational viability through outcome-independent questions about practice; the process and the dimensions are documented under Ambiflow principles. For tracking over time, a further triad has proved useful: leading process indicators, short repeated feedback from the organisation, and lagging outcome measures with an explicitly defined time horizon, which protects against reading an expected trough prematurely as failure.
SCALING
Most transformations lose their impact between pilot and full rollout
Many transformations do not fail where they started.
Pilot projects go well. New ways of working are adopted. Teams report better results. The first feedback is positive. Confidence builds in the steering group.
The real test comes afterwards.
Between a successful pilot and a changed organisation lies a challenge that gets remarkably little attention in most transformation models: scaling.
Accompany boards and executive teams through larger change initiatives and the same pattern recurs. What worked in one unit produces less effect at a second site. In the third, unexpected difficulties appear. After a few years there are convincing individual cases, while the intended transformation of the organisation as a whole has not materialised.
The cause often does not lie in the pilot.
Organisations differ more than programmes initially assume. Leadership teams, market conditions, personal constellations, historical experience and local routines all shape how new ways of working are received. What seems self-evident in one unit can meet entirely different preconditions in another.
Many pilots also benefit from conditions that cannot be reproduced later.
They receive particular attention. Extra resources are available. Experienced people are deliberately involved. Difficulties get addressed quickly. And everyone involved knows the initiative is being watched.
Under those conditions valuable insight emerges. It does not automatically produce a realistic picture of what scaling will look like.
The decisive question is therefore under which conditions a pilot works, and only secondarily whether it works.
For leadership bodies this changes the view of success. A pilot does not primarily prove that a solution works. Its real value lies in making the preconditions of its effectiveness visible.
This is where many misjudgements arise.
Positive results are often read as confirmation of the concept although they rest in part on local particularities. Critical factors go undetected. The organisation scales a solution before it has understood why the solution works at all.
A second problem compounds this.
Knowledge travels considerably less well than is usually assumed (Szulanski 1996).
In successful pilot units, teams develop more than new processes. They develop practical experience, shared routines and tacit knowledge. Much of it is never documented. It shows in how exceptions are handled, in informal agreements, or in decisions that feel obvious because they were practised together over months.
Precisely that knowledge tends to be lost in the rollout.
The receiving organisation adopts the visible elements of a concept without having gone through the same learning. On paper the solution is adopted. In practice something else comes into being.
Scaling therefore consists at its core in spreading understanding; spreading measures alone is not enough.
A further finding deserves attention. Successful scaling rarely comes from addition alone.
Every new way of working displaces existing routines. Every new priority competes with existing priorities. Every additional initiative claims attention and management capacity.
Organisations that keep adding new demands without ending old ones eventually overload their own capacity for change.
Scaling therefore demands the same discipline as transformation itself. Not only in building the new, but in deliberately letting go of the old.
For boards and executive teams a central diagnostic question follows: which conditions make the pilot successful, and which of them actually exist across the organisation?
The earlier that question is asked, the more realistic expectations about time, investment and effect become.
Scaling is thereby a leadership task in its own right and much more than a downstream delivery step.
Many transformations earn their reputation as a success in the pilot. Whether they deserve it is decided in the rollout.
Pilot effects shrink at scale for four reasons
Many transformations arrive at the same moment.
The pilot counts as a success. The results convince. The feedback is positive. In the leadership body an expectation forms that the effect can now be extended step by step across the organisation.
This is often where the hardest phase of the initiative begins.
What worked in the pilot unit frequently produces markedly less effect at scale. Research speaks of a systematic loss of effect between pilot and scale, the voltage drop (Al-Ubaydli, List and Suskind; summarised in List 2022). For leadership bodies the label matters less than the question of why the pattern recurs so reliably.
Four causes appear particularly often.
The first cause lies in the pilot itself.
Pilots often run under particular scrutiny. Experienced managers are involved. Resources are made available deliberately. Difficulties get attention quickly. At the same time there is a natural incentive to make progress visible.
This does not mean pilot results are manufactured. It does mean it is not always clear how much of the effect comes from the solution and how much from the exceptional conditions around it.
The second cause lies in the selection of pilot units.
Pilots rarely take place under average conditions. Units are often chosen because they are especially motivated, have engaged leadership or bring comparatively favourable preconditions.
For the pilot that is reasonable. For scaling it creates a risk. The organisation starts inferring the general case from an exceptional one.
Anyone who has extended a successful initiative to further sites or business units knows the effect. The average of an organisation usually resembles its best pilot unit only to a limited degree.
The third cause concerns delivery at scale.
Between pilot and rollout the conditions often change. Support intensity drops. Resources are spread more thinly. Key people are not available everywhere. Individual elements get simplified or dropped.
Which produces a paradoxical situation: the organisation believes it is rolling out the same solution although it is in fact already working with an altered version.
Many scaling problems originate exactly here.
The fourth cause lies in the system.
Organisations respond to change. What produces positive effects in one unit alters, when applied broadly, the very conditions under which the original effect arose.
A new compensation model, an additional service offering or a changed leadership routine can be convincing locally. Introduced organisation-wide, they create new dependencies, conflicts or displacement effects. The conditions under which the pilot succeeded no longer exist in the same form.
As it spreads, the solution changes the system it is meant to work in.
For boards and executive teams an important consequence follows.
The success of a pilot answers only one of two questions. It shows that an approach works under certain conditions. It does not yet answer whether those conditions also exist at scale.
A pilot should therefore test more than the solution. It should equally make the preconditions for its scaling visible.
Which elements are indispensable? Which can be adapted? Which capabilities have to be built? Which resources are genuinely required? And which assumptions rest on particularities of the pilot environment?
The earlier these questions are asked, the lower the risk of mistaking local success for organisational effectiveness.
For leadership bodies this holds one of the most important lessons of scaling research: a successful pilot is a hypothesis that still has to prove itself at scale; as proof of a transformation's success it does not yet qualify.
Knowledge travels worse than the rollout plan assumes
Even where the pilot effect is real, transferring it stays difficult, and the reasons are better researched than practice suggests. Szulanski (1996) studied the internal transfer of proven practice in large companies and found the barriers in an unexpected place. Transfer fails less through a lack of motivation in the receiving units than through problems of knowledge: through causal ambiguity, because those involved themselves only partly know why their practice works; through limited absorptive capacity on the receiving side; and through the sheer effort of building the relationship between source and recipient. Research calls this phenomenon stickiness, the tendency of knowledge to adhere to the place where it arose.
Winter and Szulanski (2001) drew a conclusion from this that runs against common practice. Successful scalers copy precisely at first: the validated core arrangement is transferred exactly, and localisation begins only once the template demonstrably works in the new unit. The common counter-strategy, letting every unit adapt from the start, regularly destroys the very mechanisms for whose sake the scaling was undertaken. The real skill lies in the prior distinction between which elements belong to the effective core and which are zones for adaptation. That distinction is an empirical task for the pilot and must not become a negotiating chip in the rollout.
Scaling holds only with subtraction and rhythm
When transformations lose their effect, the cause does not always lie in the quality of the idea, the leadership or the delivery.
Often it lies in a simpler assumption: that an organisation can absorb new things without limit.
That assumption rarely survives contact with reality.
In most companies change is designed through addition. New initiatives arrive. New processes are introduced. New requirements emerge. Each individual measure looks reasonable and is often sensible.
Over time, though, the organisation's experience changes.
Employees work to new rules while still meeting the old expectations. Managers are asked to set new priorities while carrying unchanged responsibility for existing tasks. Projects start without earlier programmes ever having been properly closed.
The result is layering rather than transformation.
Accompany leadership bodies over time and the mechanism becomes visible. The organisation looks busy, engaged and in motion. At the same time, actual change falls short of expectations. The reason is rarely rejection of the change. The number of demands simply grows faster than the ability to convert them into stable routines.
Which is why successful scaling begins with more than building the new.
It begins equally with deliberately letting go of the old.
Every new way of working raises the question of which existing one can end. Every additional priority requires a decision about which other priority loses weight. Every new structure needs room that existing structures previously occupied.
Transformation is therefore always a process of subtraction as well (Sutton and Rao 2014).
The thought looks unspectacular. In practice it is among the most demanding leadership tasks there is. Deciding on the new attracts attention. Deciding to end existing routines, programmes or habits attracts far less.
At the same time, experience shows that organisations lose their capacity for change more often through unresolved legacy than through ambitious goals.
Alongside subtraction, a second factor plays a central role: the rhythm of change.
Many transformation programmes follow an implicit ideal of continuous acceleration. As soon as an initiative shows visible effect, the wish arises to speed up its spread. Change should reach the wider organisation as fast as possible.
Organisations, however, do not run purely on the logic of programmes.
New ways of working have to be practised. Managers have to accumulate experience. Teams have to develop local solutions. Routines need time to become second nature. Recent management research accordingly describes scaling as a distinct organisational phase with a rhythm of its own (definitional harmonisation in the Journal of Business Venturing 2024).
Without those consolidation phases, change often stays at the surface.
What appears on a project plan as delivered is not yet carried in daily work. The organisation knows the new logic but has not yet internalised it.
Which is why durable transformations often follow a different pattern than their programmes suggest. Phases of change are followed by phases of stabilisation. Expansion is followed by consolidation. Learning is followed by application.
Such a rhythm has to be designed deliberately; it rarely establishes itself.
For boards and executive teams a central leadership question follows: what do we have to end so that the new can work?
That question is frequently worth more than the search for additional initiatives.
Because scaling happens where organisations have enough room to build the new and leave the old behind; maximum speed contributes little to it.
Transformation therefore demands both: the discipline of subtraction and a feel for the right rhythm.
AI EVIDENCE
AI research delivers the first causal evidence on the conditions for success
Organisational transformation has been researched for decades. Much of what executives know today about change, culture or adaptability comes from case studies, surveys and the analysis of individual companies.
That research has broadened the understanding of transformation considerably. One central question nevertheless stayed open for a long time: which conditions actually contribute to success, and which merely accompany transformations that happen to succeed?
Research on AI adoption is changing that starting position.
For the first time, studies are available that observe thousands of companies over several years while separating cause from effect methodologically; the central work links survey waves covering tens of thousands of American manufacturing plants (McElheran, Yang, Kroff and Brynjolfsson 2025, U.S. Census Bureau). For transformation research this is more than a technological special case. It is a rare opportunity to test core assumptions of the field with far greater precision.
The results deserve attention.
They first confirm an observation many leadership bodies know from experience. New technology does not produce its effect automatically.
Introducing AI does not lead directly to higher productivity. In many companies the opposite happens at first. Processes fall temporarily out of balance. Roles change. Established ways of working lose their orienting force. The organisation invests time and resources before any measurable return appears. Research describes this pattern as the J-curve: the return on a general purpose technology appears only once the complementary capital of processes, data and capability has been built, and that build-up phase shows in the figures as a trough (Brynjolfsson, Rock and Syverson 2021).
For boards and executive teams the finding matters particularly.
Technological transformation rarely runs in a straight line. Between the decision to change and its economic effect there is usually a phase in which the cost is visible and the benefit is not.
Judge that phase purely by short-term outcome figures and you risk questioning necessary adjustments too early, or conversely continuing to fund initiatives that are going nowhere.
Research shows something more important still.
The companies that benefit from AI over the long run differ less through the technology itself than through the conditions under which they introduce it.
They invest in more than tools. They invest in data, processes, capabilities and new forms of collaboration. They treat technology as part of a broader change to their organisation rather than as an isolated project.
With that, AI research confirms a finding that runs through the whole transformation literature.
Transformation rarely comes from single measures. It comes from the interplay of several changes that reinforce one another.
One further mechanism is especially instructive.
Part of the performance loss during larger transformations arises where existing leadership and management routines are abandoned before new ones are viable; in the Census study this channel explains roughly a third of the productivity losses among older, established plants. Organisations often lose their bearings through the removal of the structures that previously enabled coordination and decision-making, rather than through the change itself.
Accompany leadership bodies through technological disruption and the pattern shows up regularly. Attention goes to the new, while the significance of existing leadership and decision practice is underestimated.
Research now supplies a rare piece of empirical evidence for this.
Transformation succeeds where organisations build new capability without losing their ability to act on the way; removing existing structures as fast as possible contributes nothing to that.
One boundary condition belongs explicitly with these findings: the Census study examines industrial AI in American manufacturing. Extending it to generative AI and to knowledge work yields plausible hypotheses, and for now that is exactly what they remain.
Across the broader picture the pattern nevertheless holds. A cross-country survey of around six thousand executives (Yotzov et al. 2026, NBER) reports, three years into the generative AI wave, that productivity effects are largely absent while adoption activity remains high, whereas field experiments at the level of individual workers show clear gains within the capability frontier of the tools (Dell'Acqua et al. 2023). The detailed evidence is developed on the topic page for adaptive organisations.
For boards and executive teams this holds an important insight.
Much of the debate about AI concentrates on technology. The real challenge usually sits at the organisational level. New tools can make individuals more productive. Whether organisational benefit follows is decided in processes, roles, leadership practice and the organisation's ability to integrate the new into existing value creation.
Which is exactly why AI research is so instructive for transformation research.
It confirms what many of the findings on this page already suggest. Transformation does not work the same way under all conditions. Its success depends on context, maturity, complementary investment and how organisational tensions are handled.
Transformation is therefore neither fully plannable nor fundamentally unpredictable.
It follows conditions that are increasingly well understood.
And that is where the real value of this field of research lies: it moves the discussion from recipes for success towards testable assumptions about the reality of organisations. How an AI introduction is set up along this logic is described on the page for Ambiflow AI transformation.
SYNTHESIS
Seven findings support a conditional understanding of transformation
Discussion of organisational transformation tends to look for a simple explanation. For a model that creates orientation. For a method that reduces complexity.
Research paints a more demanding picture. The reason lies in its precision: it describes more exactly under which conditions particular assumptions hold and where their limits lie. The further research has developed, the clearer one insight has become: transformation follows no universal recipes. It follows conditions.
From the mass of studies, theories and observed practice, seven findings can be condensed that matter particularly for leadership bodies.
1. Transformation can be distinguished precisely from other change.
Not every large change is a transformation. Organisational transformation begins where the value creation logic, the identity or the fundamental interpretive frames of an organisation change. This distinction is not a matter of terminology. It determines what form of leadership and steering makes sense.
2. There is no universally correct method.
Research supports individual principles rather than finished recipes. Diagnosis, participation, orientation, early wins and the alignment of structures are among the better-evidenced factors (Stouten, Rousseau and De Cremer 2018). Which combination holds depends on context, starting position and the kind of change.
3. The organisation's responses are not a side issue.
Transformations are not carried by strategies or programmes alone. Perceived fairness, the quality of participation, the intelligibility of decisions and readiness for change all shape the course of deep change directly. What often gets labelled a soft factor is among the more robust findings in the field.
4. An organisation's capacity to absorb change is finite.
Organisations cannot process an unlimited number of changes at once. Attention, capacity to learn and leadership time are scarce resources. Transformation therefore demands prioritisation. It demands the ability to build the new while deliberately ending the old.
5. Scaling is a leadership task in its own right.
Between a successful pilot and a changed organisation lies what is often the hardest phase of a transformation. Effect does not transfer automatically. Knowledge stays bound to experience, local particularities shape outcomes, and successful approaches change the conditions of their own success as they spread.
6. Decision quality deserves attention of its own.
Transformations are steered through a multitude of decisions under uncertainty, rarely through single large ones. The quality of those decisions can be judged independently of the eventual outcome. Good leadership shows in staying able to work with uncertainty rather than avoiding it.
7. Recent AI research confirms the significance of conditions.
Technology alone produces no transformation. What proves decisive are the organisational preconditions that accompany it: data, processes, capabilities, leadership structures and the ability to integrate new possibilities into existing value creation. Research now supplies evidence for this that can separate cause from effect cleanly.
Above these seven findings stands one overarching insight.
Transformation research suffers less from a shortage of models than from a shortage of discrimination. Too often outcomes get confused with their causes. Success is explained in hindsight. Good results are attributed to a method although it remains unclear what actually produced them.
Which is why the quality of diagnosis gains importance.
Organisations that want to understand transformation seriously begin with the question of which conditions hold in their own situation, and only afterwards with the question of the right method.
The stance looks modest at first. In practice it is demanding.
It requires the willingness to observe your own organisation more closely than the latest management trends. It requires making assumptions explicit and testing them regularly. And it requires accepting that transformation is neither fully plannable nor arbitrary.
For boards and executive teams this yields an orientation that is at once simple and demanding.
Real transformation becomes more likely when an organisation understands the conditions under which change can take effect in its own context. The search for the right method alone does not achieve that.
That is where any serious engagement with organisational transformation begins. How transformind works on this diagnosis is described on the page for transformation consulting. Anyone who would first like to examine their own position with an outside view will find an accessible entry point in executive sparring: sixty minutes, no mandate and no agenda. This work does not suit every organisation; it shows its value where a leadership team is prepared to take an honest measure of its own position before it acts.
FREQUENTLY ASKED