Leadership and organizational behaviour · Module 2

Organizational design and corporate culture

How should an organization divide and coordinate its work, and what is culture, beyond the slogans, that it could be designed or changed at all?

Beyond the IB course UndergraduateMBAProfessional About 17 minutes · 2,896 words · checked 16 November 2026

From the IB course to this module

The IB course describes structures (tall and flat, centralized and decentralized, matrix, Handy's shamrock) in 2.2 and types of culture and culture clashes in 2.5. This module explains why organizations take the shapes they do, using information processing, contingency theory and Mintzberg's configurations, and treats culture as a set of shared assumptions with measurable effects. It then asks the harder question the course leaves open: how far either can be engineered.

Section 01: Conceptual foundation

Every organization solves two problems at once. It must divide work into specialized tasks, which economists since Adam Smith have seen as the source of productivity, and it must coordinate those tasks so that they add up to a product a customer will pay for. Lawrence and Lorsch (1967) called the two forces differentiation and integration. Structure, the formal pattern of roles, reporting lines and decision rights, is the organization's main answer, and every structure trades something away: specialization against responsiveness, efficiency against adaptability, control against initiative.

Formal structure never covers everything. Rules cannot anticipate every situation, so organizations also rely on what people take for granted about how things are done. That is culture: in Schein's (1985) definition, the shared basic assumptions a group has learned as it solved its problems of adapting to its environment and holding itself together, assumptions that worked well enough to be taught to newcomers as the right way to think.

The two are bound together. Structure shapes who talks to whom every day, and so which assumptions form. Culture fills the gaps between the rules, and in doing so works as a control system, often a cheaper and more powerful one than supervision (O'Reilly and Chatman, 1996). This is why redesigning the boxes on an organization chart so often disappoints: the boxes change and the assumptions do not.

Section 02: Advanced explanation

Why structures differ: information and contingency

The most useful single idea in organization design comes from Galbraith (1973, 1974): the greater the uncertainty of a task, the more information must be processed while it is being done. Rules, hierarchy and plans handle predictable work. When surprises multiply, decisions pile up at the top and the hierarchy overloads. The designer then has two families of response:

  • Reduce the need for information: add slack (buffers, longer deadlines) or create self-contained units, each with the resources for its own product or market, so fewer issues cross boundaries.
  • Increase the capacity to process it: invest in vertical information systems, or build lateral relations that let people coordinate directly, from liaison roles and task forces to full-time integrators and, at the extreme, the matrix.

Contingency research reached similar conclusions from the other side. Burns and Stalker (1961) found that British firms moving into fast-changing electronics did better with organic structures, fluid roles and lateral communication, while firms in stable markets did well with mechanistic ones. Lawrence and Lorsch (1967) found that the best performers in uncertain industries were both highly differentiated and highly integrated: specialized units that thought differently, held together by strong integrating roles.

Chandler's (1962) history of large American firms added a dynamic: as companies diversified, functional structures broke under the load and the multidivisional form followed. Structure follows strategy became a maxim, though Hall and Saias (1980) pointed out that existing structures also shape which strategies managers can see.

Mintzberg's configurations and the matrix problem

Mintzberg (1979) synthesized the research into five parts of any organization (strategic apex, middle line, operating core, technostructure of analysts, support staff) and six coordinating mechanisms (mutual adjustment, direct supervision, and the standardization of work, outputs, skills and, in later work, norms). Real organizations tend to settle into a few configurations, each built around one dominant mechanism:

  • Simple structure: direct supervision by the founder. Fast, but fragile.
  • Machine bureaucracy: standardized work in a large, stable operation. Efficient, but rigid.
  • Professional bureaucracy: standardized skills, as in hospitals and law firms. Professionals have great autonomy and resist managerial control.
  • Divisionalized form: standardized outputs, with divisions judged on performance targets.
  • Adhocracy: mutual adjustment among experts in project teams. Innovative, but costly and hard to scale.

The value of the scheme is diagnostic: many failed reorganizations impose one configuration's mechanisms on another's work, such as output targets on professionals whose quality cannot be counted.

The matrix tries to have two primary dimensions at once, such as product and country, with dual reporting. It suits firms that genuinely need both global efficiency and local responsiveness, but it slows decisions, blurs accountability and generates conflict. Bartlett and Ghoshal (1990) argued that the successful multinationals built a matrix in managers' minds, shared priorities and informal networks, rather than relying on the formal grid.

What culture is, and how it can be measured

Schein distinguished three levels. Artefacts are what an outsider sees: office layout, dress, rituals, stories. Espoused values are what people say they believe, as in mission statements. Basic assumptions are beliefs so taken for granted that nobody states them, such as whether bad news may be taken upward. Artefacts are easy to observe and hard to interpret; assumptions are what actually drive behaviour.

Martin (1992) warned that "the culture" of a large organization is often a fiction. The integration perspective sees shared values; the differentiation perspective sees subcultures, such as engineering against sales or head office against the field; the fragmentation perspective sees ambiguity and shifting alliances. All three are usually true at once.

Measurement is hard. Surveys can map culture onto typologies, such as Cameron and Quinn's competing values (clan, adhocracy, market, hierarchy), but self-reports miss what people take for granted and what they would rather not say. Researchers now also analyse text, from employee reviews to managers' language on earnings calls.

What does culture do? Sørensen (2002) found that firms with strong cultures had more reliable performance in stable industries, but the advantage disappeared in volatile ones, where strong cultures adapt slowly. Guiso, Sapienza and Zingales (2015) found that values proclaimed on corporate websites were unrelated to performance, while employees' perceptions of management's integrity were related to it. Culture is what people experience, not what the company announces.

Can culture be changed?

Because assumptions are taken for granted, they cannot simply be replaced by announcement. Schein's account of how leaders shape culture points to the levers that work: what leaders consistently pay attention to, measure and control; how they react to crises; how they allocate rewards and status; and whom they recruit, promote and remove. Structures, procedures and stated values reinforce these primary mechanisms, but cannot substitute for them.

The practical implication is to change behaviour first. When structures, incentives and routines make a new way of working rewarding, and leaders visibly act on it, new assumptions can form from experience. Values statements that contradict daily experience teach the opposite lesson, that management's words are not to be believed.

Culture also has a dark side. Strong norms that make an organization efficient can make deviance invisible: Vaughan (1996) traced how engineers and managers at NASA came to treat repeated anomalies as normal before the loss of the Challenger shuttle, a process Vaughan called the normalization of deviance. And culture clashes are a known hazard of mergers: in experiments Weber and Camerer (2003) showed that groups with different working conventions performed worse after merging than before.

Executives take culture seriously. In a large survey Graham, Grennan, Harvey and Rajgopal (2022) found that most executives believed culture matters for firm value and that their own culture was not where it should be.

Section 03: Models and frameworks

Where each comes from, how to use it, and where it stops being useful.

Mintzberg (1979)

Structural configurations

Organizations have five basic parts and coordinate through a handful of mechanisms. They tend to settle into configurations, simple, machine bureaucracy, professional bureaucracy, divisionalized and adhocracy, each built on one dominant way of coordinating.

How to use it

Identify the dominant coordinating mechanism the work actually needs, then check that structure, controls and incentives fit it. Ask which configuration a proposed reorganization assumes, and whether the work would really be coordinated that way.

Where it fails

The types are ideal forms; most large organizations are hybrids with different configurations in different units. The scheme describes fit but says little about how to move from one configuration to another, and it predates digital coordination tools that change the cost of mutual adjustment.

Henry Mintzberg (1979). The Structuring of Organizations: A Synthesis of the Research. Prentice-Hall.

Galbraith (1973)

The information-processing view of design

Task uncertainty determines how much information an organization must process during execution. Designs either reduce that need (slack, self-contained units) or raise the capacity to meet it (information systems, lateral relations, integrators, the matrix).

How to use it

Map where uncertainty enters the work and where decisions queue. Then choose the cheapest mechanism that matches the load: a liaison role or a standing team often solves a coordination problem without a full reorganization.

Where it fails

It treats organizations as information processors and says little about power, motivation or identity, which often decide whether a design works. Uncertainty is also hard to measure in advance.

Jay R. Galbraith (1973). Designing Complex Organizations. Addison-Wesley. Jay R. Galbraith (1974). Organization design: an information processing view. Interfaces, 4(3).

Schein (1985)

Levels of culture

Culture exists at three levels: visible artefacts, espoused values and taken-for-granted basic assumptions. Leaders embed it chiefly through what they attend to, reward and tolerate, and through how they react to crises.

How to use it

In a diagnosis, start with artefacts and stated values, then look for contradictions between them; contradictions point to the real assumptions. In a change programme, check the primary embedding mechanisms first: what gets measured, rewarded and punished.

Where it fails

The model is hard to test because basic assumptions are by definition unstated; diagnosis depends on the skill of the observer. It also leans towards an integration view and can underplay subcultures and the power of those who define what the culture is.

Edgar H. Schein (1985). Organizational Culture and Leadership. Jossey-Bass.

Waterman, Peters and Phillips (1980)

The 7S model

Seven elements must fit each other: strategy, structure and systems (the hard elements) and shared values, style, staff and skills (the soft ones), with shared values at the centre. A change to one element that leaves the others unaligned is likely to fail.

How to use it

Before a reorganization, describe each of the seven as it is and as it must become, and look for the elements left out of the plan: a new structure with unchanged systems and rewards is the classic gap.

Where it fails

It is a checklist, not a causal theory: it says everything must fit but not which element drives the others or how to sequence change. Its authors drew it from consulting experience rather than systematic tests.

Robert H. Waterman Jr, Thomas J. Peters and Julien R. Phillips (1980). Structure Is Not Organization. Business Horizons, vol. 23, no. 3.

Section 04: Real business applications

Real organizations appear only with dated, cited facts; illustrative examples are labelled.

Real organization, 2026-02

Unilever: from a matrix to five Business Groups (2022 to 2025)

On 25 January 2022 Unilever announced that it would move away from its matrix structure and organize around five Business Groups: Beauty & Wellbeing, Personal Care, Home Care, Nutrition and Ice Cream. Each would be fully responsible and accountable for its strategy, growth and profit globally, supported by Unilever Business Operations, which would supply technology, systems and processes, and by a lean corporate centre setting overall strategy. The change would remove around 15% of senior management roles and 5% of more junior management roles, around 1,500 roles in all. The chief executive spoke of making the company more responsive with "crystal-clear accountability for delivery".

Its 2025 results, published in February 2026, recorded the next steps: the Ice Cream business had been demerged in December 2025; separate sales organizations by Business Group had been created in the largest markets to strengthen accountability and speed, while a "One Unilever" model simplified operations in smaller markets; and the company named "a play to win culture with clear accountability" as one of three shifts for the future.

Interpretation. The design follows the information-processing logic: categories became the primary axis, scale was kept through shared operations, and the front line was split by category only where markets were big enough to carry the duplication.

Real organization, 2018-03

A regulator treats culture as a cause (FCA, 2018)

In a 2018 discussion paper the UK's Financial Conduct Authority said culture was widely accepted as a key root cause of the major conduct failings in financial services. It defined culture as the habitual behaviours and mindsets that characterize an organization. Rather than trying to assess mindsets directly, it said it would assess four drivers it could observe: a firm's purpose, its leadership, its approach to rewarding and managing people, and its governance. It stated that there was no single right culture. Essays in the collection warned that surveys may miss crucial factors because of the limits of self-report and the normalization of unethical behaviour, and that a focus on the "tone from the top" alone overlooks how complex culture is.

Interpretation. Measuring the drivers rather than culture itself is a pragmatic answer to the measurement problem, and it mirrors Schein's embedding mechanisms.

Illustrative

A consultancy that copied a start-up's squads

A professional services firm reorganizes its consultants into cross-functional squads with rotating product owners, copying a software company it admires. Within a year quality complaints rise. The firm's work depends on standardized skills and professional judgement, a professional bureaucracy in Mintzberg's terms, and the squads removed the senior review that had carried its standards. The firm keeps the squads for internal tools and restores practice leaders for client work.

Section 05: Case studies

Fictional case, simulated figures

Brightwater Software: the reorganization that worked, mostly

Brightwater Software, a fictional maker of accounting software, ran for years as a matrix: engineers reported both to product managers and to functional heads for engineering, design and quality. Customers complained that requested changes took weeks, and engineers said nobody seemed to own a decision. In 2024 the chief executive replaced the matrix with five self-contained product units, each with its own engineers, designers and testers, and cut a layer of management. A year later the board asks whether the change worked. The people team has assembled the figures below.

Brightwater Software before and after the reorganization
Measure 2023 (matrix) 2025 (product units)
Employees1,2001,250
Managers207149
Average span of control5.88.4
Management layers from chief executive to front line75
Median days from customer request to approved change4123
Decisions needing sign-off from two managers64%18%
Engineers agreeing that they know who decides (survey)38%71%
Separate tooling licences bought for the same function311
Annual staff turnover14%17%

Simulated figures. Average span of control is the number of employees other than the chief executive divided by the number of managers.

Questions

  1. Check the spans of control against the headcounts. What did the reorganization buy, and what did it cost?
  2. Use the information-processing view to explain why decisions became faster while duplicated tooling grew.
  3. Is the rise in turnover more likely an effect of structure, of culture, or of neither? What further evidence would you ask for before deciding?
  4. Propose one integrating mechanism that would recover some of the lost scale without restoring the matrix, and say what it would cost.

Section 06: Critical evaluation

What the evidence supports, where the debate is, and what the models leave out.

Structure. Contingency theory's central claim, that performance depends on the fit between structure and environment, is plausible and widely accepted, but the evidence is weaker than its reputation. Most studies are cross-sectional, fit is measured in different ways, and organizations can reach similar performance through different designs. The direction of causation is also disputed: Chandler's "structure follows strategy" has been answered by research showing that existing structures shape what managers perceive and choose. Fashionable designs such as squads, holacracy and flat structures rest mainly on case studies written by advocates, often early in the experiment.

Culture. Scholars disagree on what culture is. Smircich (1983) distinguished treating culture as something an organization has, a variable managers can adjust, from treating it as something an organization is, which cannot be managed in the same way. The early claims that strong cultures cause high performance drew on samples of successful firms selected after the fact. Better evidence supports narrower claims: strength makes performance more reliable in stable settings (Sørensen, 2002), and perceived integrity matters while proclaimed values do not (Guiso, Sapienza and Zingales, 2015).

Culture surveys. Engagement and culture surveys are useful for spotting change over time, but they are self-reports, prone to social desirability and fear, and averages hide subcultures. Once scores become targets for managers, they stop measuring what they were designed to measure. Treat a survey as one signal to be checked against behaviour: who gets promoted, what happens to people who raise bad news, and what leaders spend their time on.

Section 07: Decision exercise

A fictional situation. Decide before you read the guidance.

Integrate the start-up, or leave it alone?

Merrow Foods, a fictional packaged-food company with 6,000 employees, has bought Pip & Pod, a 90-person plant-based snacks start-up, for its fast-growing brand and its product team. Pip & Pod has an informal, founder-led culture and a two-week product cycle; Merrow's cycle is nine months, with formal stage gates. Merrow's finance team has identified £4m a year of savings from moving Pip & Pod onto Merrow's purchasing, manufacturing and sales.

  1. A. Full integration

    Fold Pip & Pod into Merrow's snacks category within six months, on Merrow's processes and pay structure. Captures all the savings quickly.

  2. B. Keep it separate

    Run Pip & Pod as an autonomous unit reporting to the chief executive, sharing only purchasing. Protects the culture but forgoes most of the savings.

  3. C. Integrate the back, protect the front

    Move purchasing, manufacturing and finance onto Merrow's systems now; keep brand and product development autonomous for two years, linked by a small integration board.

Judge the options on

  • Where the acquired value lies
  • Retention of the founders and product team
  • Speed and size of the cost savings
  • Risk of culture clash and lost innovation
  • Control and compliance risk
Read the guidance

Start with what Merrow paid for. If the value were distribution savings, full integration would be logical. But the price was for a brand and a product team whose speed comes from a different way of working, a mismatch of configurations: an adhocracy being placed inside a machine bureaucracy. Imposing Merrow's stage gates would remove exactly what made the unit valuable, and the people who carry that capability are the most likely to leave.

C is usually the strongest choice. Purchasing and manufacturing are where Merrow's scale adds value and where Pip & Pod's culture matters least, so most of the £4m can be captured there. Product and brand stay protected, and the integration board is the lateral link that keeps the two connected. B protects the culture but leaves money on the table and risks isolating the unit. A suits a deal bought for its distribution, which this was not. Whatever the choice, decide in advance which practices are non-negotiable, such as food safety and financial controls, so that autonomy does not become a compliance gap.

Section 08: Further reading

  • Henry Mintzberg (1979). The Structuring of Organizations: A Synthesis of the Research. Prentice-Hall.

    The synthesis of research on structure. Long, but the chapters on the configurations reward reading in full.

  • Jay R. Galbraith (1974). Organization design: an information processing view. Interfaces, 4(3).

    A short article stating the information-processing view of design; the clearest starting point.

  • Paul R. Lawrence and Jay W. Lorsch (1967). Organization and Environment: Managing Differentiation and Integration. Division of Research, Graduate School of Business Administration, Harvard University.

    The classic study of differentiation and integration across industries with different levels of uncertainty.

  • Edgar H. Schein (1985). Organizational Culture and Leadership. Jossey-Bass.

    The standard account of what culture is and how leaders create and change it.

  • Jesper B. Sørensen (2002). The strength of corporate culture and the reliability of firm performance. Administrative Science Quarterly, 47(1).

    Rigorous evidence on when strong cultures help and when they hinder.

  • Luigi Guiso, Paola Sapienza and Luigi Zingales (2015). The value of corporate culture. Journal of Financial Economics, 117(1).

    Separates proclaimed values from perceived integrity, with large-sample evidence.

Every work cited here is listed in the platform's sources register. Real organizations appear only with facts from the sources shown; everything else is labelled illustrative or fictional.