There is a seductive moment in almost every acquisition when the future organisation appears to become clear. The boxes have been drawn. Duplicate roles have been identified. Functions have been combined, reporting lines agreed and the names of the new leadership team inserted into the structure. The presentation may describe an integrated organisation, and legally it is. Yet very little of the work that made either organisation successful has necessarily been integrated at all. Customers still depend upon particular relationships, specialists still carry knowledge that is difficult to see from the organisation chart, decisions still travel through established networks, and teams still rely upon routines that developed over years. The chart has changed faster than the work.
This is more than a familiar complaint about reorganisations. It exposes a deeper problem in how we think about post-acquisition integration. Organisation charts describe formal accountability and authority; they do not adequately describe how value is created. Work crosses the boundaries between boxes through handovers, conversations, systems, professional relationships, judgement, informal coordination and accumulated knowledge. Some of those connections should undoubtedly change after an acquisition. Others may be precisely what the investor has paid to acquire. If leaders begin by combining structures and subsequently expect the work to adapt, they risk treating the visible organisation as though it were the organisation itself.
A better starting point is therefore deceptively simple: before redesigning the organisation, understand the work.
Standing on the shoulders of work-design research
Lynda Gratton's work on redesigning work offers a particularly useful way into this problem. Her immediate context was the disruption created by the COVID-19 pandemic and the subsequent struggle to design hybrid work, but the underlying argument is much broader. In How to Do Hybrid Right, Gratton argued that leaders should consider jobs and tasks, employee preferences, projects and workflows, and inclusion and fairness, while separating questions of place from questions of time. Her later book Redesigning Work develops a four-stage process: Understand, Reimagine, Model and Test, Act and Create.
What makes this relevant to M&A is not Gratton's particular answer about hybrid working. It is her refusal to begin with a predetermined organisational solution. COVID had disrupted assumptions about where and when work had to happen, creating an opportunity to reconsider arrangements that organisations had previously regarded as inevitable. Gratton subsequently argued that the post-pandemic transition would take years rather than quickly settling into a stable new model. The continuing argument about office attendance reinforces the larger point. The interesting organisational question is not whether remote work or office work is universally superior, but what configuration of tasks, people, interactions and context best enables particular work to be performed. Gratton's enduring contribution is therefore methodological: understand the work before deciding how it should be organised.
This places her within a much longer tradition of organisational psychology and work design rather than outside it. A century of work-design research contains several overlapping traditions, including sociotechnical systems, job characteristics, demands and control, demands and resources, and role theory. Sociotechnical thinking is especially relevant because it emerged from the proposition that the social and technical aspects of work should be designed together rather than allowing technology or formal structure to dictate how people must organise themselves. Contemporary reviews continue to emphasise joint consideration of people, technology, organisational choice and adaptability, while also acknowledging that attempts to achieve this in practice have produced mixed results.
That qualification matters. There is no magical work-design methodology that allows leaders to inspect an organisation and calculate its optimal form. Organising is a continuing act of judgement involving competing objectives. Efficiency may conflict with resilience. Standardisation may improve coordination while reducing local responsiveness. Autonomy can preserve specialist judgement while making cross-company cooperation harder. Centralisation may create visibility at the cost of slower decisions. An acquisition makes these tensions unusually visible because two functioning systems have suddenly been placed inside a common boundary. The opportunity is not merely to choose between them. It is to ask whether something better can now be designed.
You can merge the boxes overnight. You cannot merge the work overnight.
Consider what happens when two apparently similar departments are combined. From the integration team's perspective, duplication may be obvious. Both companies have finance, sales, engineering, operations, HR and marketing functions. Consolidating them promises simpler accountability and lower cost. Yet identical labels can conceal very different work. One sales team may depend upon long-standing personal relationships while another relies upon a repeatable digital process. One engineering function may develop mature products under tightly controlled procedures while another explores uncertain technologies through rapid experimentation. One regional operation may follow different practices because local regulation requires them, while another has simply accumulated inefficient habits.
The organisational question is therefore not simply whether two departments perform the same function. It is what work they actually perform, what that work depends upon, and which of those dependencies should survive the acquisition. That distinction becomes particularly important where innovation is part of the investment thesis. Puranam, Singh and Zollo's research into technology acquisitions describes a coordination-autonomy dilemma: acquirers need sufficient integration to exploit the acquired capability while preserving enough autonomy to avoid damaging continued exploration. Their findings suggest that the appropriate structural response can depend upon the developmental stage of the acquired organisation's innovation trajectory, rather than there being one universally correct level of integration.
This is an important challenge to the language of “best practice”. The acquiring company's process may indeed be better. The acquired company's process may be better. Neither may be appropriate for the organisation that is now being created. Treating integration as work design opens a third possibility: rather than deciding whose system wins, leaders can ask what system the combined work now requires.
That represents an important progression through this series. Earlier articles have been concerned with seeing and protecting value that could otherwise disappear. The Internal Market After an Acquisition examined the political system through which competing claims for authority, resources and attention are resolved. Work design turns that diagnosis into a constructive question. Instead of asking only how we prevent the politics of acquisition from destroying valuable capability, we can ask how the work itself should be redesigned so that the combined organisation becomes capable of more than either predecessor.
Start with dependencies, not departments
The real architecture of an organisation lies partly in its dependencies. A customer promise made by sales may create work for operations. An engineer's design decision may alter maintenance requirements years later. A regulatory judgement may depend upon information distributed across several specialists. A local manager may know that an apparently inefficient practice exists because of a customer expectation that has never been formally documented. A senior executive may appear to own a decision while relying heavily upon an experienced employee who knows which questions need asking before that decision can safely be made.
These relationships matter because organisational performance depends upon more than the competence of individuals. Meta-analytic evidence on team information sharing, covering 72 independent studies and 4,795 groups, found positive relationships between information sharing and team performance, cohesion, decision satisfaction and knowledge integration, although the strength of those relationships varied according to task and information conditions. Research on team cognition similarly supports the importance of shared cognitive structures for coordination and performance. DeChurch and Mesmer-Magnus' meta-analysis of 65 studies found substantial relationships between team cognition and teamwork processes, motivational states and performance, while their separate analysis of shared mental models showed that how knowledge is structured across a team matters, not merely whether individuals possess relevant knowledge.
Transactive memory adds another layer. Teams do not need everybody to know everything; they benefit from knowing, in effect, who knows what. Recent meta-analytic work reports a positive relationship between transactive memory systems and team outcomes while also showing that effect estimates vary according to measurement and context. That nuance is useful for M&A because the transaction disrupts precisely the conditions through which collective memory operates. People may remain employed while the network that made their expertise accessible is broken.
An organisation chart can show that Jane now reports to Ahmed rather than Sarah. It cannot show that six people previously knew to ask Jane whenever a particular customer made an unusual request. Nor can it guarantee that Jane knows who possesses the equivalent expertise in the acquired organisation. Combining teams therefore creates a cognitive-design problem as well as a structural one. The combined organisation has to develop new shared understandings of the task, new knowledge about where expertise resides and new routines for bringing differentiated knowledge together. This is one reason why integration can initially make a larger organisation less capable. It possesses more knowledge in aggregate while becoming temporarily worse at finding and combining it.
The earlier series article Protect What You Paid For approached this as capability continuity. Here the question moves one step further: how should the new work be designed so that valuable knowledge is not merely retained but becomes easier for the combined organisation to find, share and develop?
Aviation offers a useful comparison, but not a template
Aviation has spent decades dealing with a version of this problem because operational teams are often assembled temporarily from people who may never previously have worked together. A captain and first officer can meet shortly before a flight and still be expected to operate safely in a complex, dynamic environment. The system does not assume that interpersonal familiarity will somehow emerge in time to make coordination possible. Instead, it provides common procedures, shared terminology, explicit roles, standard briefings, checklists, communication conventions, recurrent training and clearly defined authority.
Yet the interesting feature of good Crew Resource Management is that standardisation does not remove judgement or flatten authority. The captain remains accountable, but other crew members are expected to contribute information and challenge when circumstances require it. Procedures create a shared platform from which expertise can operate; they are not intended to replace expertise. The aviation system therefore combines hierarchy with voice, standardisation with adaptation, and individual responsibility with collective cognition.
This is relevant to post-acquisition integration, but it should not be imported carelessly. A commercial organisation is not an aircraft cockpit. Tasks are less bounded, goals are more contested, feedback is slower and political interests are often more consequential. Aviation also benefits from extensive regulation, recurrent training and mature operating standards that most organisations neither possess nor would wish to reproduce. The useful lesson is narrower: effective coordination between people with differentiated expertise does not require everybody to become the same. It requires sufficient common architecture for their differences to become usable.
That distinction should influence how we think about integration. The objective is not necessarily to create identical practices throughout the combined organisation. It is to establish sufficient shared understanding of intent, interfaces, decision rights, information and escalation that differentiated expertise can coordinate reliably. The published Observatory article 10 Ways High-Performing Teams Create Shared Reality develops this wider principle: shared reality is not unanimity, but a sufficiently shared understanding of what is happening to enable coordinated action while preserving the capacity to challenge and update that understanding.
Psychological safety belongs in the design, not on the poster
The social conditions surrounding the new work matter as much as its formal architecture. Psychological safety has become one of the more fashionable concepts in management, sometimes diluted into a general aspiration that people should feel comfortable at work. The research tradition is more specific. Frazier and colleagues' meta-analysis aggregated 136 independent samples representing more than 22,000 individuals and nearly 5,000 groups, finding empirical support for psychological safety as part of a wider network of antecedents and outcomes in workplace behaviour. In an integration context, its relevance is particularly practical: will people disclose uncertainty, challenge a questionable assumption, admit that the new process is not working or explain why a local variation exists?
This cannot be solved by telling acquired employees that their voices matter. Article 5 showed why the political context matters. If the acquiring company's executives control appointments, resources and future careers, an invitation to challenge does not occur in a socially neutral environment. Employees observe whose proposals survive, whose expertise is deferred to, what happens to people who question integration decisions and whether information that contradicts the preferred narrative changes anything. The design of meetings, escalation routes, decision processes and review mechanisms can therefore either make useful challenge easier or make silence rational.
This is where organisational psychology and organisational politics meet. Formal authority is necessary because integration requires decisions. But authority also affects whose knowledge is treated as credible. A senior executive can inadvertently turn an assumption into organisational reality simply by stating it before others have contributed. A headquarters team can describe a local practice as resistance before understanding why it exists. Equally, an acquired team can use specialist knowledge as a shield against legitimate scrutiny. Psychological safety should not mean that expertise is beyond challenge; it should mean that expertise and challenge can meet without status deciding the answer in advance. The goal is not a comfortable integration. It is an organisation capable of discovering when it is wrong.
This also connects to the published field guide Decision Rights During the First 100 Days. Formal authority, delegated boundaries and escalation routes provide part of the architecture. Work design adds the equally important question of whether the people holding authority receive the knowledge and challenge necessary to exercise it well.
Gratton's four stages become an integration discipline
Gratton's four-stage approach becomes especially powerful when applied at this point. The first stage, Understand, means resisting the temptation to begin with the target organisation chart. Instead, leaders examine the jobs, tasks, workflows, networks and human preferences through which performance currently occurs. In an acquisition we would extend this further to include critical knowledge, customer relationships, decision dependencies, informal authority and local context. The purpose is not to preserve everything that exists. It is to understand what exists before deciding which elements are dispensable.
The second stage, Reimagine, is where the acquisition should begin to earn its strategic premium. If integration consists solely of moving the acquired company onto the buyer's systems and eliminating duplicated cost, the organisation may capture efficiencies without creating much new capability. Reimagining asks what becomes possible because previously separate resources can now be combined. Could local customer knowledge be joined with greater technological capability? Could a mature operating system support an entrepreneurial product team without suffocating it? Could two specialist networks create a stronger knowledge system? Could employees gain career pathways and development opportunities that neither organisation could previously offer? These questions move integration beyond absorption towards organisational creation.
Model and Test, Gratton's third stage, may be the one most frequently neglected in M&A. Integration programmes are often under pressure to demonstrate pace and certainty, which encourages irreversible decisions before their consequences are well understood. Yet many organisational choices are testable. A cross-company team can be piloted before the entire function is reorganised. New decision rights can be trialled against a defined class of decisions. A common customer process can be tested in one region. A local practice can be temporarily retained while evidence is gathered about whether its claimed benefits are real. The organisation does not need to treat every uncertainty as an excuse for delay, but neither does it need to manufacture certainty merely because the integration plan contains a deadline.
Finally, Act and Create means embedding what has been learned into the operating system: structures, technology, roles, measures, training, leadership expectations and decision rights. This prevents experimentation from becoming perpetual pilot activity. The organisation eventually has to choose. But the choice is now informed by a more deliberate understanding of the work and by evidence generated through testing rather than solely by hierarchy or inherited preference.
There is an important political implication here. Modelling and testing changes the nature of disagreement. Instead of two groups arguing indefinitely that their approach is superior, they can make assumptions explicit and ask what evidence would distinguish between them. Not every dispute can be settled experimentally, and organisational interests will not disappear simply because data are available. But shifting some integration disagreements from whose view wins? towards what would we need to observe to know? is itself an improvement in organisational judgement.
Protect, Change, Combine, Localise, Learn
The work-design perspective also gives greater precision to the five integration judgements emerging through this series. Protect applies where existing work contains capability that remains important to the investment thesis: a specialist network, customer relationship, effective routine, trusted brand, decision capability or body of tacit knowledge. Protection should not mean freezing it permanently. It means avoiding accidental destruction while the organisation understands what creates the value.
Change applies where the existing design constrains future performance. An acquisition may reveal duplicated activity, weak controls, founder dependency, poor management, fragmented systems or processes that have survived through habit rather than usefulness. The argument for understanding work before changing it is not an argument for organisational conservatism. Sometimes the most respectful response to acquired employees is to fix problems they have been living with for years.
Combine is where much of the acquisition's potential lies. The purpose is not merely to put similar people under one manager but to identify complementary capabilities that become more valuable when connected. Combining customer access with technical expertise, operational discipline with entrepreneurial experimentation, or one organisation's knowledge with another's scale can create capability that did not previously exist. This is integration in its strongest sense.
Localise acknowledges that variation is sometimes functional. Different regulatory environments, customer expectations, labour markets, cultures and operating conditions can justify different practices. But localisation needs evidence just as standardisation does. “We have always done it this way” is not sufficient, and neither is “headquarters wants one process”. The question is what variation the work genuinely requires.
Finally, Learn creates a legitimate category for decisions where the evidence is not yet sufficient. This may be the most important addition because integration rhetoric tends to reward decisiveness. Leaders can feel pressure to choose between competing models even when neither side knows enough to justify certainty. Learning does not mean postponing decisions indefinitely. It means specifying what is uncertain, what will temporarily happen, what evidence will be collected and when the question will be reconsidered.
Together, these categories transform integration from a binary contest between preservation and change into a continuing exercise in organisational judgement.
A Post-Acquisition Work Design Map
We can make this practical. Before finalising the structure of a strategically important part of the combined organisation, leadership can create a Post-Acquisition Work Design Map. The map begins not with reporting lines but with six dimensions: Purpose, Work, Knowledge, Coordination, Authority and Context. For each, the integration team asks what outcome must be produced, which tasks and workflows create it, where the necessary expertise resides, which dependencies require coordination, where decisions should sit and which contextual differences genuinely require variation.
Those six dimensions are then worked through Gratton's four stages: Understand → Reimagine → Model and Test → Act and Create. At each stage, the integration team can apply the five judgements developed in this series: Protect → Change → Combine → Localise → Learn. This creates a deliberately layered framework rather than another acronym. Gratton provides the redesign process; organisational psychology helps us understand the work and teams; organisational politics reminds us that redesign redistributes interests and authority; the five integration judgements force explicit choices about what happens to existing capability.
This should not be presented as a new theory of organisation design. Its intellectual ancestry is clear. Gratton provides a contemporary redesign methodology; sociotechnical systems emphasise the joint design of people and technology; work-design research provides a much wider evidence base; team cognition and transactive memory explain aspects of coordination and distributed knowledge; M&A research exposes the autonomy-coordination dilemma; organisational politics reminds us that apparently technical design decisions distribute power. The potential contribution is their synthesis around a particular practical problem: how do we redesign work after an acquisition without assuming that the transaction itself tells us what the new organisation should look like?
That is a useful contribution precisely because it does not pretend to have invented the component ideas.
The integration team is temporary. The organisation it creates is not.
There is one final danger. Integration teams themselves can become temporary shadow organisations. They acquire decision rights, create workstreams, resolve conflicts and push actions through an organisation under exceptional conditions. This can be necessary. But eventually the programme office disappears, consultants leave and senior executives turn their attention elsewhere. The permanent organisation inherits whatever routines, interfaces and unresolved dependencies the integration process has created.
A successful integration should therefore be judged partly by whether ordinary teams can operate effectively after extraordinary integration governance has been removed. Do people know who knows what? Can consequential decisions be made at an appropriate level? Can local knowledge challenge central assumptions? Are interfaces between functions understood? Can new employees learn how the organisation really works without finding the right veteran? When circumstances change, can teams adapt without escalating everything back to the executives who designed the integration?
These are work-design questions, but they are also questions of organisational judgement. A capable organisation does not merely have talented individuals or an elegant structure. It enables people with different expertise, interests and perspectives to establish enough shared reality to coordinate, decide and learn under uncertainty.
That is why the post-acquisition organisation chart should be an output of integration thinking, not its starting point.
You can merge the boxes overnight. You cannot merge the work overnight.
The more ambitious objective is not to make two organisations look like one. It is to design a system in which their combined people, knowledge, relationships and technology can become capable of something neither organisation could reliably achieve alone.
And that is where integration begins to become value creation.
Continue the series
This is Article 6 in Buying the Business You Think You Are Buying. The preceding article, The Internal Market After an Acquisition, examined how acquisition redistributes power, status, resources and attention. Earlier articles explore what can disappear after completion, why individually sensible integration decisions can cumulatively change the organisation, and how to protect critical capability while reducing dependency. The next article turns the lens outward again: customers did not agree to the acquisition, so how does the combined organisation continue earning their trust while its own work, people and identity are changing?
.svg)
.svg)
