Question: A short Business Review/Article Analysis for "The Emergence and Evolution of the Multidimensional Organization." (overview/analysis of major points)

The Emergence and Evolution of the Multidimensional Organization J. Strikwerda J.W. Stoelhorst \"In terms of its impact, not just on economic activity, but also on human life as a whole, the multidivisional organizational design must rank as one of the major innovations of the last century.\"John Roberts1 T he multidivisional, multi-unit, or M-form, is widely acknowledged as the most successful organization form of the twentieth century.2 Firms that employ the M-form organize their activities in separate business units and delegate control over the resources needed to create economic value to the managers of these units. This organization form is widespread, is central to the \"theory in use\" of managers, and serves as the basis of most accounting systems. However, the organization of productive activities in many contemporary rms violates the principle that is central to the M-form: that business units are self-contained. The quest for synergies that has been high on the corporate agenda since the late 1980s has resulted in the widespread adoption of corporate account management, shared service centers, and matrix organizations. As a result, most business units now depend at least in part on resources that are controlled by other units. This raises fundamental questions about the status of the M-form in contemporary rms. Questioning the status of the M-form is not merely a theoretical fancy, but is high on the agenda of managers as well. In this article, we report on research that was commissioned by the Foundation for Management Studies, a Dutch organization of management executives. These practical men and women shared a fundamental uneasiness about structuring their organizations. On the one hand, many of them experienced problems with the M-form: high employee costs, internal battles over resources, lack of standardization, lack of cooperation, and loss of market opportunities. On the other hand, they did not CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 11 The Emergence and Evolution of the Multidimensional Organization see any viable alternatives to the multi-unit organization form. The need to exploit synergies across business units was widespread, but it was unclear which organizational designs are most appropriate to achieve this. This led to a research project to explore the ways in which leading Dutch organizations, including subsidiaries of foreign multinationals, have adapted the M-form to better exploit synergies across business units. As we expected, the results of the study vividly illustrate the fundamental tension between the need for contemporary rms to exploit synergies and their need for clear accountability. However, an additional and unexpected nding was that a number of rms in the study have evolved an organizational form that signals a new way of resolving this tension. These rms are organized around multiple dimensions (e.g., region, product, and account) and simultaneJ. Strikwerda is a Professor of Organization and Change at the Amsterdam Business ously hold different managers accountable for School, University of Amsterdam and performance on these dimensions. This mulDirector of the Nolan Norton Institute at Zeist, tidimensional organization form is based on Netherlands. different principles than the M-form, the most J.W. Stoelhorst is an Assistant Professor of notable of which is that resources and market Strategy and Organization at the Amsterdam Business School, University of Amsterdam. opportunities are organized separately, so that unit managers are deliberately made dependent on each other to achieve their objectives. The multidimensional organization also differs from the matrix organization. In particular, it avoids the situation where employees have two bosses. This is made possible by a different way of organizing information and by a planning and control process in which the customerrather than any one of the dimensions for which managers are held accountable (e.g., country or product line)is seen as the main prot center. First, however, a discussion of the M-form and its inherent limitations is in order. This will clarify why changes in the economic context in the second half of the twentieth century have put a strain on the way large businesses are typically organized. The signicance of the multidimensional organization is best understood against the backdrop of the evolution from a resource-centric industrial economy that was focused on exploiting tangible physical resources, to a customer-centric service economy that is focused on exploiting intangible knowledge resources. Multidimensional organizations can be interpreted as attempts to implement a team-based approach to economic value creation that seems particularly well adapted to exploiting the distributed nature of knowledge that is arguably the dening characteristic of modern economies. The Rise and Fall of the M-Form Economic historian Alfred Chandler famously documented the emergence of multidivisional organizations in the rst half of the twentieth century.3 Economist Oliver Williamson labeled this type of organization the M-form to distinguish it from the U-form that it replaced.4 In the U-form, or unitary form, 12 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization the rm is a single prot center that is organized along functional lines. The large industrial rms that Chandler described realized that a unitary organization stood in the way of their growth strategies. Instead of using functional departments such as production and sales as the main dimension along which to organize, these rms broke up into separate divisions, each of which was itself a U-form. A \"corporate parent\" acting as headquarters coordinated the divisions. This organization design is based on the following four principles: 6 The rm is organized in separate business unitsEach division targets a specic market. The specic scope of the divisions can be dened on the basis of such criteria as geography, product, customer, or distribution channel. The divisions operate as strategic business units, meaning that they can pursue a competitive strategy that is geared to the specic market on which they focus. Each business unit is a prot center, so that, in contrast to the U-form, the rm now consists of multiple prot centers. 6 Business unit managers are accountable for creating economic valueThe corporate parent holds business unit managers accountable for creating economic value. This accountability goes hand in hand with the delegation of authority. Within the business scope dened for each unit, and apart from nancing its operations, the authority to make the necessary strategic, tactical, and operational trade-offs to achieve business objectives is delegated to the business unit managers. 6 Resources are allocated unequivocally to business unitsTo be able to create economic value, business units control all the resources they need to pursue a focused competitive strategy, with possible exceptions such as research and development funds. This control allows business units to respond to market requirements with a minimum of coordination with other business units or corporate headquarters. 6 The task of the corporate parent is to add value to the activities of the business unitsHeadquarters denes the scope of the business units, allocates resources, and coordinates activities where necessary. Each business unit is an investment project in itself and the task of the corporate parent is to support the business units in creating economic value. The turnover of the rm is the sum of turnovers of the business units, corrected for internal deliveries. The income of the rm is the sum of the incomes of the business units, minus the costs of the corporate parent. These organizing principles have been applied widely in the United States and Europe, and successfully so.5 There are two reasons for the M-form's popularity and success: 6 By creating an internal capital market the M-form stimulates entrepreneurship For the corporate parent to coordinate the activities of the different business units, it is in principle sufcient to dene the business scope of each unit by specifying its target market.6 After the scope of the business units is dened, corporate headquarters essentially manages an internal capital market to allocate nancial resources to the business units. In addition, corporate headquarters may exercise control by providing overall strategic CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 13 The Emergence and Evolution of the Multidimensional Organization direction: approving nancial and non-nancial objectives; controlling performance against objectives; appointing, assessing, remunerating, and dismissing managers; and imposing corporate policies. However, despite these possible additional roles for headquarters, the main principle of the M-form is that business unit managers are responsible for creating economic value and control the resources that allow them to do so. By organizing resources on the basis of markets served and delegating control of these resources to business unit managers, rms that use the M-form can better respond to market opportunities. The M-form stimulates entrepreneurship and creates opportunities for individuals to develop into general managers without the need to provide capital. 6 The M-form offers a simple way to exploit synergiesIn terms of management accounting, the M-form is a simple organization that is nevertheless able to exploit some synergies across business units. The M-form is simple in the sense that accounting and control follow one dimension: the performance of the business units. The efciency of the rm as a whole can be assessed through its break-up value: the value of the rm must be higher than the sum of the values of the business units. This is achieved by exploiting synergies, provided these synergies do not breach the quasiautonomy of the business unit managers. Such synergies can take the form of nancial synergies and economies of scope. Williamson made a further distinction between the H-form, or nancial holding, in which the corporate parent only pursues nancial synergies, and the M-form, in which the corporate parent pursues both nancial synergies and economies of scope.7 Financial synergies may result from spreading the risks of investment across business units or reducing the costs of capital of the rm. The sources of economies of scope include corporate R&D and management development programs that benet multiple business units. Despite its success during much of the twentieth century, by the late 1980s the M-form was under attack. The critique of the M-form focused on the added value of the corporate parent to its business units.8 From the perspective of outside investors, corporate headquarters adds costs that an investor could avoid by directly investing in the business units. On this view, it is clear that the overhead costs of a corporate layer of management should be offset by the value that the corporate parent adds to the activities of its business units. If the parent is able to increase the ability of its business units to create economic value, the value of the rm as a whole will be higher than the sum of the values of its business units. However, in many cases it could be demonstrated that the overall value of the rm was in fact lower than its break-up value. In other words, rather than creating shareholder value, many parents were destroying it. The limited success of many corporate parents in adding value to their business units raised some fundamental questions about the M-form that both managers and theorists have been struggling with ever since the issue was raised some twenty years ago. It may be clear that the only source of added value is the ability of the parent to exploit synergies of some sort. For much of the twentieth 14 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization century, the relative efciency of the internal capital markets of the H-form and M-form allowed rms to reap nancial synergies. One of the main causes of capital market inefciencies is information asymmetry between shareholders and managers. Outside investors may not have access to all the inside information that is needed to assess the true value of investment opportunities. In an internal capital market, this problem is reduced because headquarters has the power to conduct nancial and strategic audits of the business units. However, by the 1980s, external capital markets had become much more efcient and there was little room left to add value through nancial portfolio management alone. With the potential for nancial synergies much reduced, the H-form became much less popular and M-form rms increasingly focused on exploiting economies of scale or scope. However, in doing so, they also ran into the limitations of the multi-unit design. The need to exploit synergies across business units has led to a variety of responses. On the market side, key account management systems are a common solution to exploit customer synergies. On the resource side, shared service centers are a common solution to exploit synergies in production, delivery, and a variety of (support) functions. Over the years, many rms have experimented with different types of matrix organizations. What all these responses have in common is that they reduce the independence of business units. This, in turn, goes against the very idea of the M-form. At its heart, the M-form assumes that each of the rm's customers only needs to deal with one business unit, and that each of the rm's business units can control all resources needed to serve its market. It is only on these two assumptions that business units can truly operate independently. These assumptions are violated as soon as there are opportunities for cross-selling or system integration across the products of the different business units, complementarities among the resources of the different business units, or economies of scale in the use of resources across business units. In other words, there is a crucial trade-off involved in choosing between the unequivocal resource allocation and lines of authority that have led to the success of the M-form, on the one hand, and the creation and the exploitation of the kind of synergies that product and nancial markets increasingly require, on the other. In Search of Alternatives for the M-Form The tension between the need to create synergies and the organizing principles of the M-form explains why the Dutch Foundation for Management Studies commissioned a study into the status of the M-form among large rms in the Netherlands (see the Appendix, \"About the Research\"). Interviews at 36 large Dutch organizations, including a number of subsidiaries of non-Dutch multinationals and a number of non-prot organizations, led to ve main conclusions about the status of the M-form. CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 15 The Emergence and Evolution of the Multidimensional Organization The M-form Still Dominates the Corporate Landscape In 29 of the 36 rms interviewed, the internal organization was based on one dimension. These rms were organized around units that were dened in terms of either regions, products, markets, product-market combinations, or in some cases distribution channels (e.g., intermediaries in the insurance industry). An interesting case among these 29 rms was the \"split business chain\" employed by Philips Medical Systems, in which development and manufacturing are organized globally in divisions based on products, while marketing, sales, and services are organized on the basis of regions. However, with the exception of Philips Medical Systems and the seven multidimensional rms identied below, the dominant way of organizing rms in the sample was on the basis of the principle underlying the M-form: around units that are dened in terms of one specic dimension. The M-form Is Central to the \"Theory in Use\" of Executives A second conclusion is that the traditional multi-unit organization is central to the \"theory in use\" of executives.9 Executives' comments on organization were typically framed in terms of the organizing principles that underlie the M-form. Clearly dened accountabilities and performance targets at the level of business units are seen as the norm against which organizational solutions are to be judged. As one respondent remarked, \"When I was appointed CEO of this rm, the organization was a complete mess. There was only one way to clear this up: to organize it strictly in business units and divisions, without compromise.\" Another illustrative comment by a CEO was: \"My rst philosophy is to decentralize until the bitter end, which should result in clearly identiable entities for entrepreneurship. My second philosophy is that entrepreneurship is by far more important than achieving synergies. My third philosophy is that if the market changes, the organization needs to be adapted.\" That the M-form dominates the thinking of executives about organization is reinforced by the fact that the multi-unit form is also the basis of the accounting systems of most companies. The Matrix Organization Is a Negative Mental Anchor In contrast to their frequent recourse to the principles of the M-form when discussing their rm's organization, executives had nothing positive to say about the organizing principles of the matrix organization. The interviews made it very clear that the concept of a matrix organization has very negative connotations. This is despite the fact that all respondents acknowledged that a rm typically cannot be managed on only one dimension. A remark that was made, in slightly different terms, in multiple interviews, was the following: \"You cannot run a company like this along a single dimension, you need to manage along multiple dimensionsbut one thing is for sure, I never again want to work with a matrix organization.\" Executives associate the matrix organization with unclear responsibilities, a lack of accountability, and political battles over resources, resulting in risk-averse behavior and loss of market share. 16 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization On the basis of the interviews, it seems that negative experiences with matrix organizations even prevent top executives from asking the crucial question that is central to recent scholarship on designing complex organizations \"are there ways to make the matrix organization work?\"10 Virtually No Business Unit Is Self-Contained Although the M-form still dominates both organization charts and executives' mental maps, at the same time, the interviews showed that virtually no business unit is self-contained. In contrast to the ideas underlying the M-form, all rms are organized on the basis of business units that depend on resources outside the unit to achieve their objectives, albeit in varying degrees. This is in large part the result of the use of synergy mechanisms. Both shared service centers (26 out of 36 organizations) and corporate account management (15 out of 36 organizations) are widespread. In other words, while the M-form may still dominate the thinking about organization, the actual practice of organizing has taken rms away from the underlying logic of the M-form. Mental Anchoring to the M-form Creates Problems When Implementing Synergy Mechanisms A number of rmssuch as ARCADIS, an engineering rm with worldwide operationswere wrestling with the organization of corporate account management. While ARCADIS has historically operated very successfully on the basis of independently operating business units, its corporate account management program is one of its most successful growth platforms. A corporate account is typically a customer who is served by multiple business units, so that corporate account management cuts across business units. The question that rms like ARCADIS face is whether or not to grant corporate account managers prot-and-loss responsibility. Project driven rms such as construction companies face a similar dilemma with respect to project management. In a rm such as Van Hattum & Blankenvoort (a division of one of the largest construction companies in the Netherlands, the VolkerWessels group), large projects are typically the prot drivers of the rm, yet there is also a need to monitor the protability of the business units that not only deliver goods and services to the projects, but also to their own external customers. One of the problems with respect to giving corporate account managers or project managers prot-andloss responsibility is opposition by business unit managers. They typically perceive such responsibilities as reducing their own status, power, and autonomy. In fact, in ve of the rms interviewed, the mental anchoring to the M-form resulted in a failure to successfully implement account management or project management. Together, these ve conclusions conrm the reality of the tension that executives face when organizing their rms. Most rms seem to be \"stuck in the middle\" between the principle of self-contained units that is central to the M-form, on the one hand, and the need to create synergies across business units, on the other. However, our research also unexpectedly turned up a number of CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 17 The Emergence and Evolution of the Multidimensional Organization TABLE 1. The Multidimensional Firms from the Sample Firm Number of Dimensions ABN AMRO 4 (+1) Regions, Global Clients, Market Segments, Products, (Support Functions) Ahold (Albert Heijn Company) 8 Time, Place, Formula, Category, Customer's Loyalty Card, Receipt, Regions, Branch/Store ASML 2 (+1) Products, Accounts, (Support Functions) IBM 4 (+1) Product/Solution, Regions, Accounts, Distribution Channels, (Support Functions) Microsoft 4 Products, Regions, Applications, Market Segments PricewaterhouseCoopers 2 (+1) Industries, Professional Services, (Support Functions) Van Hattum & Blankevoort 2 Business Units, Projects Dimensions rms that seem to have found a new way to balance the demands for synergies with clear accountability and control mechanisms. Multidimensional Organizations Seven of the rms that were interviewed create and exploit synergies across their business units by organizing along multiple dimensions, and ve of these rms were studied in more detail (see Table 1). One of these rms is PricewaterhouseCoopers Netherlands (PwC), the professional service rm for accounting, tax, and management consulting. This rm is organized on the basis of three dimensions: industries (including key accounts), professional services, and support functions such as HR, IT, and facility management. This means that, in addition to operating shared service centers for the support functions, PwC has made different managers responsible for pursuing market opportunities and for managing resources. The \"market managers\" of industries and key accounts have top-line responsibility. They are accountable for turnover, market position, and customer retention. However, these managers control very few resources. To reach their targets, they depend on the \"resource managers\" of the professional service units. In addition to controlling most of the resources, these resource managers also have bottom-line responsibility. The crux of PWC's organization is that pursuing market opportunities is organized separately from managing resources. In doing so, the rm knowingly violates the organizing principle that is at the heart of the traditional M-form: that resources are unequivocally assigned to unit managers so that they can use these resources to pursue market opportunities. What is interesting is that PwC's violation of this principle does not result in wheeling and dealing among 18 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization the market managers and resources managers. Instead, the inherent conict between developing new market opportunities and an efcient utilization of resources is deliberately brought to the table of PwC's executive board. To fruitfully exploit the tension between innovation and efciency, the task of the board is twofold. On the one hand, it must reduce risk-averse satiscing behavior among the resource managers by confronting them with the market opportunities identied by the market managers. On the other hand, it must confront any overly optimistic judgments about market opportunities among the market managers. The example of PwC highlights three of the six main characteristics of multidimensional organizations that we identied. Multidimensional organizations: simultaneously report their performance on two or more dimensions at multiple levels in the organization; simultaneously hold different managers accountable for the contribution of their dimension to the rm's overall performance; and organize resources in such a way that the managers who are responsible for the different dimensions are dependent on each other for the resources they need to achieve their performance objectives. What is essential is that in multidimensional organizations, the reconciliation of grasping market opportunities and using resource efciently is a corporate issue that is visible to all. Trade-offs are made in light of their effect on overall rm performance. A possibly paradigmatic example that illustrates in more detail how multidimensional organizations operate is IBM. When Gerstner took over as CEO in the early nineties, IBM was a product-centric rm with a focus on hardware products such as mainframes. Its internal organization was based on a geographical model: the key positions outside the United States were the country managers. Today, IBM offers products, services, and solutions and is organized around four dimensions: regions, products and solutions, industries and accounts, and distribution channels. Contrary to his predecessor Akers, who intended to split IBM up into a number of autonomous companies, Gerstner implemented a strategy aimed at offering integrated services, including servicing a number of global key accounts. Palmisano, the current CEO, has extended this strategy with his concept of the \"Globally Integrated Enterprise.\"11 To operate as an integrated enterprise, IBM simultaneously holds different managers accountable for the results of products, accounts, regions, and distribution channels. Product managers are accountable for turnover, margin, market share, and market penetration of specic products or solutions. Account managers are accountable for client satisfaction (which is measured twice a year by an independent agency), turnover, and margin for specic accounts. Account managers have staff for industry expertise and architecture design, but are otherwise dependent on staff and expertise from the product groups. Region managers are accountable for turnover, margin, and market penetration for specic products, but which products will be offered in a region is beyond the decision authority of the region manager. Distribution channel managers are accountable for turnover, return on sales, and customer satisfaction for their channel. Product, account, region, and distribution channel managers all depend on shared CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 19 The Emergence and Evolution of the Multidimensional Organization services organized in \"Globally Integrated Support Functions,\" including nance, HR, legal, communications, and sales support. A crucial building block of IBM's ability to operate as an integrated enterprise is its management information system. Before 1994, in keeping with IBM's geographical organization, IBM Finance was decentralized. Management information systems were geared to the needs of the local business units, cost allocations were subject to gaming among business units, and consolidation of the necessary information for nancial reports was tedious. Today, all IBM's transactions are recorded according to one set of common data denitions and processes and they are consolidated in one global general ledger. This has created a trusted central source of accounting information that can be simultaneously reported in terms of the multiple dimensions along which IBM is organized. In addition, IBM operates a CRM system that enables all managers to consult the database for orders and leads and species their contribution to customer protability. For instance, a distribution channel manager has access to information about the percentage of total sales to individual customers through her specic channel, and thus about her contribution to IBM's overall prot with this customer. IBM's management information systems are the basis of a management process that is designed to facilitate cooperation between the managers who are accountable for results on each of the different dimensions. Interestingly, the main prot centers in IBM's organization are not the units that bear responsibility for one of the four dimensions for which individual managers are held accountable. The main prot center is the customer and the primary task of all of the unit managers is to optimize IBM's position with its customers. To balance customer demands and the efcient use of resources and to be able to respond to tactical buying behavior of customers that attempt to increase their bargaining position vis--vis different IBM units, product and account managers confer each month. This is done on the basis of management information provided by IBM Finance from its accounting and CRM systems. All managers within IBM receive the same information so that horizontal information asymmetries (across dimensions) and vertical information asymmetries (across layers in the hierarchy) are eliminated. For new leads, product managers and industry consultants jointly write a business case. Depending on the situation, more weight may be given to account management or product management, but the criterion to decide on the right balance is always the contribution to the overall performance of IBM. If product and account managers cannot agree, the matter is escalated to region managers, or in exceptional cases to corporate headquarters. In addition to the three characteristics of the multidimensional form specied above, the IBM case illustrates three more features that are common to the seven rms in our sample that we identied as multidimensional organizations. Multidimensional Firms Have Multidimensional Market Positions The primary reason for organizing rms along multiple dimensions is that multidimensional rms have identied multiple dimensions that are critical to 20 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization their market position. Decision rights are allocated in line with the importance of these dimensions in the market. IBM's multidimensional organization is the result of its philosophy that it should be able to conduct business with its customers in the way that these customers prefer. Some customers prefer to do business through product-based sellers, while others prefer industry-based sales teams, solutions-based sales teams, intermediaries, or direct channels such as call centers and the Internet. As a result, IBM has dened its market in terms of no less than six segments that have signicant overlap (global accounts, regional accounts, midsized customers, industry-specic solutions, discrete products, and distribution channels). While various commentators initially criticized IBM for the complexity of its market denition and suggested that it would need to concentrate on only one dimension, this criticism was silenced by IBM's success in serving the market along these different dimensions.12 Multidimensional Firms Focus on the Customer as the Prot Center The customer-centric nature of multidimensional rms is further enhanced by the practice of treating customers as prot centers. In fact, although all the managers that carry responsibility for a particular performance dimension typically have prot responsibility, multidimensional rms see customers as the primary prot center. Multidimensional rms create economic value creation by pursuing market strategies on the basis of integrated product-and-service offerings that maximize customer protability. An example of this approach is how ABN-AMRO allocates individual customers to specic unit managers. ABN-AMRO is organized along four dimensions: regions, market segments, products, and global clients. Prot is reported for each of these dimensions, but also for each customer. This is made possible by a worldwide CRM system, in which all customers (corporate accounts, SME, or retail) are recorded down to the level of daily business and protability. On the basis of the specic prole of the customer, quarterly decisions are made to allocate each customer to the unit manager that is best placed to optimize the protability of the customer. For instance, an SME rm would typically be allocated to the SME manager in the rm's country of residence. However, if the owner of such a rm decides to sell his business, the market manager for private equity clients will typically be much better placed to protably serve this customer. In the past, it was left to the initiative of the customer to become a private equity client with ABN AMRO. Now, on the basis of the information collected by the account manager of the client (in this case the SME manager) and recorded in the CRM system, the quarterly meeting on customer reviews decides to move the customer into the custody of the private equity manager. The SME manager is rewarded by a referral fee, and the private equity manager takes on the responsibility for cross-selling products and services other than his own private equity service. CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 21 The Emergence and Evolution of the Multidimensional Organization Multidimensional Firms Eliminate Information Asymmetries and Transfer Pricing In multidimensional rms, corporate headquarters owns the transaction data and customer data. External and internal transaction data are recorded on multiple characteristics in one general ledger and are shared across the rm. The general ledger allows for consolidation along multiple dimensions as well as for reports at the level of individual customers. By operating one general ledger, multidimensional rms create a single trusted source for performance data, and by making these data available to all, multidimensional rms eliminate information asymmetries. Also, because the general ledger records all external and internal transactions, transfer prices between units and the attendant negative effect of double marginalization on the performance of the rm are eliminated. The crucial importance of information systems in allowing rms to operate as multidimensional organizations is illustrated by one of the rms in our sample that is moving towards the multidimensional form. This multinational company, which operates four divisions that are active in around sixty countries, has been running a global data system for some twenty years. The system reports along the product-driven divisional organization as well as the legal dimension of its organization (the operating companies within the divisions, as well as the national subsidiaries that act as holdings for the local sales organizations). When the rm changed its strategy to include global key account management, it discovered that the global data system could not report the necessary information. Sales to key accounts were locked up in local databases that could only report consolidated nancial gures per product division, which was, after all, the main dimension along which planning and control had historically proceeded. When the rm also decided to establish a project dimension that cuts across its four divisions to offer integrated services to the hotel market, it ran into further difculties, as the product divisions were hesitant to make the necessary investments and transfer prices undermined the protability of the projects. The challenge that the rm is currently facing is threefold: to allow the global data system to report protability on key accounts and projects, to give the key account managers and project managers prot-and-loss responsibility, and to use the global general ledger to eliminate transfer prices. The Evolution of the Multidimensional Form The ways in which multi-business rms can be organized to reap the benets of synergies across their business units has long been a central topic in the management and strategy literature. As a theoretical concept, the idea of a multidimensional organization is not entirely new. In fact, the idea that the Mform should exploit economies of scale and scope through nancial synergies, corporate management development, or R&D programs led to a line-and-staff organization that can already be seen as multidimensional, albeit in a very limited way. In the line-and-staff organization, functional managers not only report to the general managers of their business unit, but also to corporate functional 22 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization managers, so that management information needs to be made available along both the line and the staff dimensions. However, the planning and control processes in the M-form are strictly one dimensional, with the business units as the only prot centers. In the late 1970s, there was some theoretical interest in the potential of more explicitly organizing multi-business rms along several dimensions in a number of publications that were concerned with the changing nature of multinational corporations (MNCs).13 These publications are from a period in which European multinationals were transforming their country-based international operations, and U.S. multinationals their international sales divisions, into transnational organizations. The typical result of these transformations was a matrix organization, rather than a full-edged multidimensional design. As an empirical phenomenon, the multidimensional organization as we describe it is new. While it certainly has its precursors in well-designed matrix organizations, such as the much celebrated global matrix that ABB operated in the 1990s, or in so-called \"network organizations,\" it has a number of crucial features that signal a new phase in the evolution of multidimensional organization designs.14 In Table 2 we use the dimensions of Galbraith's \"star model\" for organization design to summarize the main differences between the traditional M-form, the typical two-dimensional matrix organization, and the multidimensional form.15 The table highlights the differences between the M-form and the multidimensional form and also shows that the matrix organization is in many ways an intermediate form that may share more features with either the M-form or the multidimensional form, depending on how well it is designed. In fact, one could easily conceive of a matrix organization as a two-dimensional organization, and a multidimensional organization such as IBM could conversely be seen as a very complex matrix organization. However, there are also three important differences between the typical matrix organization and the organization of the multidimensional rms in our sample. Strategy In keeping with Chandler's dictum \"structure follows strategy,\" the emergence of the multidimensional organization is best understood against the backdrop of the shift from the physical production model (that was central to the industrial revolution) to the knowledge exploitation model (that is central to a post-industrial service and experience economy). Multidimensional rms have fully acknowledged the age-old marketing wisdom that their protability ultimately depends on their ability to create value for customers. They have also acknowledged that the sources of their ability to create this value are shifting from tangible physical assets and codied knowledge to the tacit personal knowledge of creative knowledge workers. The result is strategies that are customer-centric instead of resource-centric and aimed at knowledge exploitation rather than physical production. CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 23 The Emergence and Evolution of the Multidimensional Organization TABLE 2. A Comparison of Different Organization Forms M-form Strategy Organization Planning and Control Process Matrix Form Multidimensional Form Physical Production Model (Intermediate) Knowledge Exploitation Model Resource Centric (Intermediate) Customer Centric Portfolio of Units (Intermediate) Integrated Firm Single Dimension Multiple Dimensions Multiple Dimensions One Boss Two Bosses One Boss Autonomy (Intermediate) Teamwork Focus on Tangible Resources (Intermediate) Focus on Intangible Resources (Intermediate) Resource Mobilization Resource Allocation Single Dimension Multiple Dimensions at Unit Level Multiple Dimensions at All Levels Transaction Data Owned by Units Transaction Data Owned by Regions Transaction Data in Trusted General Ledger Unit Primary Prot Center Product or Region Primary Prot Center Customer Primary Prot Center Rewards Individual Performance Targets (Intermediate) Team and Individual Performance Targets People Unit Oriented (Intermediate) Firm Oriented Hierarchical (Intermediate) Cooperative Motivated by Control of Resources (Intermediate) Motivated by Contribution to the Success of the Firm Organization Customer-centric strategies based on knowledge resources demand teamwork. Perhaps the main difference between the M-form and the multidimensional organization is that the executive board no longer sees its task in terms of portfolio management, but in terms of running an integrated rm. This is why the efcient use of resources and the pursuit of market opportunities are organized separately, so that managers become dependent on each other. The rm is no longer seen in terms of one dominant dimension along which resources are allocated to quasi-autonomous managers, but in terms of multiple dimensions that cooperate toward the common goal of creating value for customers. This unlocks scarce knowledge resources from self-contained business units, so that they can be used across the rm. Planning and Control Process The multidimensional organizations in the research sample have all dened the customer as the primary prot center in their accounting systems, and have translated this choice into their management control processes.16 This stands in contrast to most rms, where the logic of the M-form is hardwired 24 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization into the accounting systems, and performance measurement and control typically proceed along one dimension, taking the business unit or division as prot center. More often than not, this one dimension is a poor proxy for the rm's position with customers. A necessary condition for operating a multidimensional organization is that performance data is recorded and reported over multiple dimensions and at all levels of the organization. Only a central accounting system that is seen as a trusted source of nancial data can eliminate information asymmetry and the need for transfer prices between units, both of which are essential for running an integrated customer-centric rm. Trade-offs between dimensions are visible to all involved and a clear planning and control process makes sure that these trade-offs are made with an eye to improving the development and exploitation of knowledge resources across the rm and optimizing the rm's position with customers. Rewards The essence of the multidimensional organization is that it turns on the principles of \"team play.\" This means that all players know the game and their roles in this game, share information, cooperate towards a common goal, and are rewarded on the basis of their contribution towards realizing this common goal. To be able to reward the collective effort of team play, remuneration is primarily based on people's contribution to the common goal. This stands in contrast to the remuneration systems that dominated in the traditional M-form and that reward managers in proportion to their position in the vertical hierarchy. People Multidimensional rms select managers for their motivation to make a visible contribution to the overall performance of the rm. Organizing a rm on the principle of teamwork calls for another type of manager than the one that typically thrives in the M-form. Whereas the traditional business unit manager is likely to be extrinsically motivated by strong nancial incentives and the status conferred by controlling resources, the emphasis on teamwork in the multidimensional organization calls for managers that are intrinsically motivated by their contribution toward a common goal. Multidimensional rms put information systems in place that disrupt the monopoly on information that managers could use to control their subordinates in the traditional Weberian bureaucracy. Interestingly, as a female manager at IBM remarked, this development may well benet women because they are known to be more likely to share information and to solve problems collectively than men. The hiring and promotion practices of a rm thus have an important role to play in making a multidimensional organization work. While the differences between the M-form and the multidimensional organization may be clear, the differences between the matrix organization and the multidimensional form are subtler. While a poorly designed matrix organization resembles the M-form without the attendant advantages of clear accountability, a well-designed matrix organization may in many ways resemble a CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 25 The Emergence and Evolution of the Multidimensional Organization multidimensional organization. However, there are nevertheless three important differences between a multidimensional design and the matrix. The rst difference is in terms of organization. In a matrix organization, the \"node\" on the organization chart where the dimensions meet is the employee, who reports to two bosses that each have their own objectives. In a multidimensional organization, the \"node\" in the organization is no longer the employee, but the customer. In multidimensional organizations, most employees are in fact employed in specic units and report to a single manager. Rather than sharing employees, the managers who are responsible for specic dimensions (say, product and region) share customers. The second difference is in terms of the planning and control process. Rather than dening one of the dimensions as the main prot center, multidimensional organizations are primarily oriented on the protability of customers. The protability on the different dimensions is seen as a derivative of customer protability, and although each of the managers of the different dimensions typically has some form of prot responsibility, the prot on his dimension is made subservient to overall protability with customers. The third difference is that to be able to avoid the two bosses of the traditional matrix, the multidimensional organization employs information systems that can simultaneously report performance on the different dimensions at all levels in the organization, down to individual customers. These systems allow the multidimensional organization to eliminate information asymmetries and transfer pricing and to truly operate as an integrated customer-centric rm. In hindsight, the multidimensional rms in the research suggest that the root of the problems that many rms encountered with the matrix organization was its focus on authority and power, whereas the solution to making an organizational design along multiple dimensions work lies in a focus on the rm's joint goals with customers and information systems that allow rms to operate along multiple dimensions without the ambiguities that are inherent in the two-boss system. Conclusion \"In the knowledge society, managers must prepare to abandon everything they know.\"Peter F. Drucker17 For most of the twentieth century, the organizing principles of the Mform served rms well. However, contemporary economic conditions increasingly call for the creation and exploitation of synergies; and this is not something for which the M-form was conceived, or can be adapted easily. In their search for alternatives to the M-form, many rms in our sample have experimented with and subsequently discarded the matrix organization. Instead, the most common organization design is the M-form combined with synergy mechanisms such as corporate account management and shared service centers. This particular combination creates tensions because these synergy mechanisms go against the basic design principles of the M-form. However, a limited but substantial 26 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization number of rms have evolved towards multidimensional designs that allow them to better exploit the intangible resources that are increasingly becoming the main source of competitive advantage in contemporary economies. The multidimensional organization is an alternative to the M-form and the matrix organization, and it is explicitly based on design principles that facilitate the creation of synergies across units to serve increasingly fragmented markets. An important question is if the multidimensional organization is efcient. Ultimately, this question can only be answered by observing the future nancial performance of multidimensional organizations.18 However, the multidimensional form does seem to be particularly well adapted to the economic logic of a knowledge society. Knowledge-driven production calls for rms that are able to integrate distributed and tacit knowledge. Because no single individual holds all the knowledge that is required to produce the complex products and services of most contemporary rms, both the exploration and exploitation of productive knowledge requires teamwork. Seen in this light, the emergence of the multidimensional organization can be understood as a way to facilitate the teamwork that is needed to successfully create and exploit productive knowledge. Our ndings show that rms have evolved towards a multidimensional design as a result of the need to simultaneously optimize the proactive exploitation of market opportunities and the efcient exploitation of resources. On the market side, the main rationale for adopting a multidimensional organization is that many markets are no longer one-dimensional. Customer preferences have become much more fragmented and as a result products and services are increasingly sold through multiple distribution channels. The retail market that Ahold faces may be seen as an extreme case. Individual customers are likely to shop in different Albert Heijn stores (Ahold' s main brand in the Netherlands): a traditional medium-sized neighborhood Albert Heijn supermarket for utility shopping, a small \"Albert Heijn ToGo\" shop at a railway station for convenience products, or an \"Albert Heijn XL\" for leisurely shopping at the weekend. This means that the traditional organization design in which individual stores were the main prot centers can no longer serve as an accurate proxy for the company's position with customers. In fact, Ahold has realized that individual customers may have different sets of preferences depending on the time of day, the day of the week, or the location where they shop, so that in some sense even individual customers further fragment into distinct consumer behavior proles that may be triggered at different moments. On this view, Ahold's ultimate challenge is to increase the company's protability for different \"clouds of consumer behavior.\" The eight dimensions along which Ahold's planning and control processes proceed are an attempt to do just that. Another reason for the multidimensional nature of markets is that industries are increasingly networked. As a consequence, the locus of innovation has shifted from the individual company to the network in which it operates. In the case of IBM, the integration of components into a system may be a task that the customer wants to perform in-house, it may be a task that the customer wants to outsource to an independent third party, or it may be a task that the customer CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 27 The Emergence and Evolution of the Multidimensional Organization wants IBM to fulll. That IBM is organized along four dimensions reects the fact that it wants to serve all three types of customers and illustrates Ashby's \"Law of Requisite Variety\": to control a market, a rm's internal organization should at least match the complexity of the external environment. On the resource side, the central reason for adopting a multidimensional organization is that it allows rms to make better use of personal knowledge resources. This both increases the rm's ability to increase economies of scope and makes the rm more attractive for knowledge workers. The individual expertise of knowledge workers is increasingly becoming the most valuable resource of rms, and by organizing market opportunities and resources separately, individual expertise is no longer locked-up in specic business units.19 Instead, it can be applied in the products and services for the customers who most value the expertise. This is not just an advantage for the rm, but also for the knowledge worker, whose personal market is increased. This, in turn, should give multidimensional organizations a competitive advantage in the market for knowledge workers compared to the M-form. As Drucker already emphasized, in the knowledge economy, \"organizations have to market membership as much as they market productsand perhaps more.\"20 Knowledge workers, especially the more creative kind, are known to pursue as large a personal market for their knowledge as possible, and on this aspect the multidimensional organization clearly outperforms the M-form.21 It is, of course, one thing to conclude that the multidimensional organization has advantages for knowledge-driven rms that operate in complex, interrelated markets, but quite another to make a multidimensional organization work. This is where the main source of corporate added value lies. In a multidimensional organization, the role of headquarters is not merely to manage a portfolio of investments, but to create the conditions that make it possible for the rm to operate as an integrated team. The rst step towards creating these conditions is to overcome the mental barriers that are the legacy of the M-form. Both the rm's top executives and, equally importantly, its business unit managers will have to give up on some deeply engrained notions about organization design. Instead of unequivocally allocating resources, the interdependence of units is explicitly designed into the rm. The second step is to reorganize the rm's information systems, planning processes, and reward schemes. The crucial enablers for a successful multidimensional organization are systems and processes that eliminate information asymmetries and allow employees to jointly focus on creating economic value for customers. APPENDIX About the Research Interviews were conducted at 36 organizations. They are listed below, with the seven multidimensional rms highlighted in bold. Half of the interviews were with the CEO of the rm, while the other half were with the CFO 28 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization or the controller. In addition to assessing the status of the multi-unit organization as well as the use of account management and shared service centers, the interview protocol focused on questions such as: \"Which alternatives to the multi-unit organization are used?\" and \"Which trade-offs are made in deciding between familiar and new organization forms?\" The question that led to the unexpected discovery of the multidimensional organization form was \"Along which axis is accountability for turnover and prot organized in your organization?\" After the interviews, a number of multidimensional rms were selected for detailed case studies (ABN-AMRO, AholdAlbert Heijn, IBM, PwC, and Van Hattum & Blankevoort). These case studies were co-authored with representatives of the rms and reported back to the rms before being authorized for publication. Firm Industry 1. ABN-AMRO Financial Services 2. Eureko - Achmea Corporate Account Management Prot / Non-Prot X P Insurance X P 3. AholdAlbert Heijn Food Retail X P 4. Akzo-Nobel Chemicals X P 5. ARCADIS Engineering (X) X P 6. ASML Hi Tech Equipment X X P 7. Ballast Nedam Construction P 8. Bollegraaf Recycling Equipment Engineering P 9. Carante Groep Care X NP 10. Cordaris Financial Services X NP 11. Royal Cosun Food 12. DSM Chemicals 13. Essent Energy 14. Evean Groep Care 15. IBMNetherlands IT Equipment and Services 16. ING Financial Services 17. KLM Cargo Air Transport 18. KPN Telecommunication CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 X Shared Service Centers P X X P Pa X NP X P X P X X P X X P X SUMMER 2009 CMR.BERKELEY.EDU 29 The Emergence and Evolution of the Multidimensional Organization Firm a. b. c. d. Industry 19. MicrosoftNetherlands Software and IT Services 20. Ministry of Interior 21. Ministry of Defense Corporate Account Management X Shared Service Centers Prot / Non-Prot X P Government X NP Government X NP c 22. Nationale Nederlanden Insurance P 23. PGGM Financial Services NP 24. Philips Electronics Electronics X X P 25. PricewaterhouseCoopers Professional Services X X P 26. Rabobank Financial Services X X Pb 27. Rodamco Europe Real Estate X 28. Sabic Europe Chemicals X X P 29. Sanoma Publishing X X P 30. Shell Oil and Energy X P 31. SNS Reaal Financial Services X P 32. TNT-Post Postal Services X P 33. Unilever Food X P 34. UWV Social Security NP 35. Van Hattem & Blankevoort Constructiond P 36. Vitens Public Water Supply Pa X P Shareholders are public bodies Cooperative with a statutory injunction on distributing prot to its members Division of ING Division of VolkerWessels Notes 1. John Roberts, The Modern Firm: Organizational Design for Performance and Growth (Oxford: Oxford University Press, 2004), p. 2. 2. Oliver E. Williamson, The Economic Institutions of Capitalism (New York, NY: The Free Press, 1985), p. 279. 3. Alfred D. Chandler, Strategy and Structure: Chapters in the History of American Enterprise (Cambridge, MA; MIT Press, 1962). 4. Oliver E. Williamson, Markets and Hierarchies (New York, NY: The Free Press, 1975). 5. Neil Fligstein, \"The Spread of the Multidivisional Form Among Large Firms, 1919-1979,\" American Sociological Review, 50 (1985): 377-391. 6. Alfred P. Sloan, My Years with General Motors (Hammondsworth, UK: Penguin Books, [1962] 1986). 30 UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU The Emergence and Evolution of the Multidimensional Organization 7. Williamson (1975), op. cit. 8. Michael Goold and Andrew Campbell, Strategies and Styles: The Role of the Centre in Managing Diversied Corporations (Oxford: Basil Blackwell, 1987); Michael E. Porter, \"From Competitive Advantage to Corporate Strategy,\" Harvard Business Review, 65/3 (May/June 1987): 43-59. 9. For the concept of \"theory in use,\" see Donald A. Schn, The Reective Practitioner: How Professionals Think in Action (Aldershot, UK: Arena, 1991). 10. Jay R. Galbraith, Designing Matrix Organizations That Actually Work (San Francisco, CA: JosseyBass, 2009). 11. Samuel J. Palmisano, \"The Globally Integrated Enterprise,\" Foreign Affairs, 85/3 (May/June 2006): 127-136. 12. Jay R. Galbraith, Designing the Customer-Centric Organization: A Guide to Strategy, Structure, and Process (San Francisco, CA: Jossey-Bass, 2005): 112. 13. Russell L. Ackoff, \"Towards Flexible Organizations: A Multidimensional Design,\" Omega, 5/6 (1977): 649-662; Christopher A. Bartlett, \"How Multinational Organizations Evolve,\" The Journal of Business Strategy, 3/1 (Summer 1982): 20-32; William C. Coggin, \"How the Multidimensional Structure Works at Dow Corning,\" Harvard Business Review, 52/1 (January/February 1974): 54-65; C.K. Prahalad, \"The Concept and Potential of Multidimensional Organizations,\" in F. Stevens, ed., Managing Managers (Eindhoven: N.V. Philips' Gloeilampenfabrieken, 1980), pp. 159-176; C.K. Prahalad and Yves, L. Doz, \"Strategic Reorientation in the Multidimensional Organization,\" Graduate School of Business Administration, The University of Michigan, Working Paper No. 195, 1979. 14. For well-designed matrix organizations, see Galbraith (2009), op. cit. For ABB, see Christopher A. Bartlett and Sumantra Ghoshal, \"Beyond the M-Form: Toward a Managerial Theory of the Firm,\" Strategic Management Journal, 14 (Special Winter Issue 1993): 23-46. For more in-depth and critical discussions, see Christian Berggren, \"Building a Truly Global Organization? ABB and the Problems of Integrating a Multi-Domestic Enterprise,\" Scandinavian Journal of Management, 12/2 (1996): 123-137; Winfried Ruigrok, Leona Achtenhagen, Mathias Wagner, and Johannes Ruegg-Sturm, \"ABB: Beyond the Global Matrix Towards the Network Multidivisional Organization,\" in Andrew Pettigrew and Evelyn Fenton, eds., Processes and Practices of New Forms of Organizing (London and Thousand Oaks, CA: Sage, 2000). For network organizations, see Michael Goold and Andrew Campbell, \"Structured Networks: Towards the Well-Designed Matrix,\" Long Range Planning, 36/5 (October 2003): 427-439. 15. Jay R. Galbraith, Organization Design (Reading, MA: Addison Wesley, 1977). See also Galbraith (2009), op. cit. Note that we focus on the planning and control process, which is essential to the operation of the multidimensional organizations we studied. We do not discuss operational processes, which differ widely across the rms. 16. See also Galbraith (2005), op. cit. 17. Peter F. Drucker, \"The New Society of Organizations,\" Harvard Business Review, 70/5 (September/October 1992): 95-104. 18. Among the seven multidimensional rms in the sample, six rms are performing well. ABN AMRO is the exception. Under pressure from activist shareholders that were dissatised by the performance of its shares, it has recently been taken over by a consortium of three competitors and broken up. This illustrates that the multidimensional organization, like any form of organization, is certainly no panacea. 19. C.K. Prahalad and Gary Hamel, \"The Core Competence of the Corporation,\" Harvard Business Review, 68/3 (May/June 1990): 79-91; Bartlett and Ghoshal, op. cit. 20. Drucker (1992), op. cit., p. 100. 21. Richard Florida, The Rise of the Creative Class (New York, NY: Basic Books, 2004); Sherwin Rosen, Markets and Diversity (Cambridge, MA: Harvard University Press, 2004). CALIFORNIA MANAGEMENT REVIEW VOL. 51, NO. 4 SUMMER 2009 CMR.BERKELEY.EDU 31