Designing Matrix Organizations that Actually Work E-Book

Jay R. Galbraith

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Sprache: Englisch

Seitenzahl: 321

Veröffentlichungsjahr: 2008

Beschreibung

Organization structures do not fail, says Jay Galbraith, but management fails at implementing them correctly. This is why, he explains, the idea that the matrix does not work still exists today, even among people who should know better. But the matrix has become a necessary form of organization in today's business environment. Companies now know that if they have multiple product lines, do business in multiple countries, and serve many customer segments through a variety of channels, there is no way they can avoid some kind of a matrix structure and the question most are asking is "How do we learn how to operate the matrix effectively?" In Designing Matrix Organizations That Actually Work, Galbraith answers this and other questions as he shows how to make a matrix work effectively.

Leseprobe

Table of Contents

Title Page

Copyright Page

Dedication

Preface

Introduction - MATRIX ORGANIZATIONS

What Is a Matrix?

What Are the Origins of the Matrix?

What Happened?

The Star Model

Implications of the Star Model

Part One - SIMPLE MATRIX ORGANIZATIONS

Chapter 1 - SIMPLE MATRIX STRUCTURES

Two-Dimensional Structures

Pharmaceutical R&D Lab Example

Summary

Chapter 2 - THE TWO-HAT MODEL

What Is the Two-Hat Model?

Examples of Two-Hat Structures

Summary

Chapter 3 - THE BATON PASS MODEL

The Consumer Goods Model

The Pharmaceutical Model

Summary

Chapter 4 - THE MATRIX WITHIN A MATRIX

Design Challenges of the Matrix Within a Matrix

Matrix Within a Matrix at the Corporate Level

Mars Pet Food Example

Summary

Chapter 5 - BALANCING POWER AND DEFINING ROLES

Designing Power Bases

Roles and Responsibilities

Summary

Part Two - COMPLEX MATRIX STRUCTURES

Chapter 6 - THE THREE-DIMENSIONAL MATRIX

International Strategy

The Geography-Dominant Matrix

The Balanced Matrix

The Business-Dominant Matrix

Differentiated Structures

Other Three-Dimensional Models

Summary

Chapter 7 - MORE COMPLEX MATRIX STRUCTURES

Global Account Teams

The Front-Back Hybrid Model

Summary

Chapter 8 - THE IBM STRUCTURE

The IBM Front-Back Hybrid

More Complexity?

Summary

Part Three - COMPLETING THE STAR MODEL

Chapter 9 - COMMUNICATION IN THE MATRIX

Informal Communication

Formal Communication

Summary

Chapter 10 - PLANNING AND COORDINATION PROCESSES

Goal Alignment, Dispute Resolution, and Coordination Mechanisms

Summary

Chapter 11 - PLANNING PROCESSES IN THE COMPLEX MATRIX

What About Complex Matrix Designs?

Get the System in a Room

Online Processes

Summary

Chapter 12 - HUMAN RESOURCES POLICIES

Human Capital

Social Capital

Summary

Chapter 13 - LEADERSHIP IN A MATRIX ORGANIZATION

Seeing That Conflicts Are Resolved

Managing the Top Team

Balancing Power

Summary

Chapter 14 - IMPLEMENTING A MATRIX

Using the Star Model

Building Capabilities

Summary

Chapter 15 - A SYNOPSIS OF MATRIX CAPABILITIES

Epilogue

References

About the Author

Index

Copyright © 2009 by Jay R. Galbraith

Published by Jossey-Bass A Wiley Imprint 989 Market Street, San Francisco, CA 94103-1741—www.josseybass.com

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400, fax [phone], or on the Web at www.copyright.com. Requests to the publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, 201-748-6011, fax [phone], or online at www.wiley.com/go/permissions.

Screen Beans art is copyrighted by A Bit Better Corporation and used here with permission. Screen Beans is a registered trademark of A Bit Better Corporation.

Readers should be aware that Internet Web sites offered as citations and/or sources for further information may have changed or disappeared between the time this was written and when it is read.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Jossey-Bass books and products are available through most bookstores. To contact Jossey-Bass directly call our Customer Care Department within the U.S. at 800-956-7739, outside the U.S. at 317-572-3986, or fax [phone].

Jossey-Bass also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.

Library of Congress Cataloging-in-Publication Data

Galbraith, Jay R. Designing matrix organizations that actually work : how IBM, Procter & Gamble, and others design for success/Jay R. Galbraith. p. cm.—(The Jossey-Bass business & management series) Includes bibliographical references and index.

eISBN : 978-0-470-45054-3

1. Matrix organization. 2. Corporate culture. I. Title. HD58.5.G35 2009 658.4’ 02—dc22 2008021043

This book is dedicated to my wife, Sasha.It is her willingness and ability to read and improvemy writing that helps me immeasurably. Dedicating the bookto her is my way of showing my heartfelt appreciation.

Preface

On several occasions throughout my career, I have considered writing a book on matrix organizations. Each time I decided not to use “matrix” in the title and wrote about organization design or global organization instead. In the 1970s, matrix was too trendy. I thought that once the backlash set in, the book would stop selling. Unfortunately, I was correct. By the 1980s, it was a common belief that matrix structures do not work. Under these conditions, no one would buy the book to begin with. But throughout the 1980s and 1990s, companies continued to introduce matrix or matrix-like structures. To avoid questions from their bosses, the adopters used other names, such as “multidimensional structures” or even the old “line and staff” nomenclature. I remember a project that I had at Kodak. McKinsey recommended a new strategy and a “shared resource” structure. It was a matrix structure with a new label. My learning from these clients was that matrix is an appropriate organization for many business situations. And when experienced managers face these business situations, they adopt a matrix organization, whatever it is called, to perform the business activities.

Matching a matrix organization to the appropriate situation was part of the challenge; getting it to work was a bigger one. Most managers drew the organization charts, debated where the dotted and solid lines would go, and then announced the new matrix structure to their organization. And in most cases, the structure did not work. (In this book, we will see why.) At one point I was counting the successes and failures. As I remember now, it was one in four that were successes. That is, 75 percent of the attempts failed to meet their objectives, and many actually caused more problems than they solved. It was no wonder that the idea that “matrix does not work” took hold. But I still had data points showing that 25 percent of the attempts were successful. Then one day I was calling on a client who was about to abandon the matrix. At lunch I was talking with a disappointed manager. He said the company had not changed the performance appraisal system or the planning process as I and others had advised. He then said, “You know, it’s not that matrix is a failure here. It’s that we’ve failed at matrix.” The statement captured many of my experiences. I had seen many instances where matrix was implemented poorly and incompletely. Those same managers who saw that matrix was the right structure were not able or willing to do the hard implementation work to make it a success. So one of my purposes in writing this book is to articulate the practices of those 25 percent who were successful.

This idea that matrix does not work still exists today, even among people who should know better (Bryan and Joyce, 2005). Organization structures do not fail; managements fail at implementing them correctly. For example, in the mid-1990s, Sun Microsystems reorganized from a functional structure to an autonomous divisional structure. Management wanted to create miniature Suns to restore the entrepreneurial spirit of the company’s start-up days. So they created divisions for high-end servers, desktop computers, printers, software, services, and so on. There were nine divisions in all. This configuration led them to call the structure Sun and the nine planets. Within a few years, Sun discovered the negatives of this structure. Management found that they had nine different compensation plans, nine accounting systems, nine IT systems, nine sales forces calling on the same customers, and so on. They had nine of everything and escalating overhead. It was almost impossible to move talent from one division to the next. Sun then abandoned the autonomous divisional structure and moved to one similar to its old one. Does this mean that autonomous divisional structures do not work? Of course not. The divisional structure serves United Technologies’ and General Electric’s diversified strategies very well. Sun simply failed to implement the structure in its specific situation. What was the structure the company moved to? You guessed it, the matrix. Sun increased the strength of HR, finance, and marketing across a reduced number of divisions, forming the typical divisional-functional matrix. It formed a single sales force and matrixed product sales specialists between the global sales organization and the product divisions. So one purpose of this book is to dispel the idea that matrix doesn’t work. Certainly there are failures, but they are failures of management.

In the 1990s, I was asked by Ed Schein and Dick Beckhardt to revise my 1973 book. I took the opportunity to write about designing all types of lateral or horizontal organizational forms, including matrix forms. We changed the title to Competing with Flexible Lateral Organizations. Whether it was the unappealing title or the decline of the OD Series I am not sure, but the book never caught on.

My current interest began while I was on the faculty at the International Institute for Management Development (IMD) in Lausanne, Switzerland. Each year, IMD ran the CEO Roundtable for the CEOs from those companies that were partners of IMD. In 1998, my colleague Ulrich Steger surveyed these CEOs for the top challenges that they were facing. From their responses he compiled what he called the CEOs’ Agenda. Some of the challenges were as expected: globalization, intense price competition, and managing constant change. But one was unexpected: it was termed “managing organizational complexity.” My colleagues asked me to handle this topic on the agenda. IMD then set up about ten interviews with CEOs and their leadership teams for my preparation. What they told me was that the customer was the new source of complexity. The ABB team said that DaimlerChrysler pointed out to them that there were thirty-seven sales forces from ABB calling on the company. DaimlerChrysler was getting thirty-seven different levels of service from these sales forces, whose members did not know each other. The company asked ABB for a single strong interface. Nokia Networks was facing a consolidating customer base as telecom operators were acquiring each other. IBM’s customers were requesting IT solutions rather than buying stand-alone products and integrating them themselves. In all cases, these companies were creating global account teams and gathering them into customer segment units. The complexity came from the fact that they were already struggling to manage through three-dimensional matrix structures of global business units, regions and countries, and global functions. And now customers were requiring a fourth dimension—customers—to be added to these global structures.

I have written about designing these complex organizations in Designing the Global Corporation (2000) and Designing the Customer-Centric Organization (2005). My intention was to address the complexity that derived from the new customer dimension. But as a result of this focus, I did not address the matrix organization specifically, nor did I title either of the books Matrix Organization Design.

But today I am sensing a renewed interest in matrix organizations. I conduct a lot of workshops for the in-house organization design groups that are appearing in an increasing number of companies. I usually tailor the sessions to the interests of the company. When I ask for the topics the managers want addressed, matrix is always on the list. Usually it takes the form of “How do we make a matrix work?” My interpretation of these requests is that we may have turned the corner on matrix. It is no longer a structure to be avoided. Instead it is a necessary form of organization in today’s business environment. So the issue today is “How do we learn how to operate the matrix effectively?” Thus it seems to me that a book focusing on how to make a matrix work may finally be a timely contribution.

Indeed, there are still a lot of managers who do not know how to make a matrix work. The following quote is from the (now former) CEO of ABB, Fred Kindle, who responded to a case writer’s question about what he thought of ABB’s matrix under prior CEOs:

“For me, [the old ABB] was like the famous movie—The Matrix. Total confusion. Nobody knows what is going on, and I don’t like that. Most important is clarity and to some extent simplicity. I am absolutely willing to trade in a 100% perfect—academically perfect—but complex organization for only [a] 70% perfect organization that is simple and easy to understand.”

He [the CEO] recalled an instance when he asked his team whether they knew to whom the country manager of Denmark reported. The answer was correct—Zürich—but nobody could name the executive. (Pucik and Zalan, 2007, p. 11)

Here we have a top executive espousing a typical view that matrix doesn’t work because it creates confusion and lack of clarity. Now, if there is confusion and lack of clarity about roles and responsibilities in an organization, my question is, “Whose fault is that?” In my view, it is the leader’s fault because he or she does not define the roles and processes through which decisions get made. In Chapter Five, I review responsibility charts, which are used in successful matrix implementations. Every consulting firm now promotes its own similar tool. Confusion is not inherent in a matrix. It is, however, often found in poorly implemented matrix organizations. I am sure that if you went to Procter & Gamble, IBM, Nokia, or Toyota, the head of Western Europe in these companies would know the executive to whom the country manager of Denmark reports. It is clear that this former ABB CEO has no idea how an effective matrix works. I am sure that he has never seen or experienced one. It is for managers like this that I wrote this book.

The other astonishing concept expressed in the quote is that the CEO is willing to accept 70 percent of the ideal to get simplicity. Matrix organizations are all about achieving two or more diametrically opposed goals and doing them both well at the same time. The goal is to execute a strategy that is both global and local. Or it is striving to deliver new products on compressed time frames and to achieve functional excellence. I cannot conceive of General Electric or Procter & Gamble or Toyota settling for 70 percent. Leadership in a matrix is all about mastering the requisite complexity to attain the 100 percent. The quote serves to highlight what it takes to lead a successful matrix organization. Again, one needs to understand what a matrix really is and what the challenges are in order to execute it successfully.

This book is based on my experiences of studying and consulting with matrix organizations for forty years. I started in summer 1967 studying the commercial airplane division of Boeing. Since then, I have worked with computer companies, consumer goods companies, commercial banks, investment banks, telecoms, retailers, semiconductor companies, aerospace firms, Internet companies, and hospitals, among others. I have worked with companies from most countries in Europe, Japan, Latin America, Israel, Indonesia, and the Middle East, as well as North America. Each industry, each country, and each company has its own unique features, but there are some core concepts that make up the content of this book. These concepts are taken from past books, my current workshop notes, and from my most recent experiences.

The book follows the Star Model that I have used in all my books. In the Introduction, I present the definition of matrix organizations, why they are chosen, and why there were failures initially. One of the reasons is that the changes were structure-only. Few companies introduced a complete organization design. The successful ones did implement a complete design. For me a complete design follows the Star Model, with aligned changes in structure, processes, rewards, and people practices. So an explanation of the Star Model completes the introductory chapter.

The book is written in three parts. Part One presents the simple two-dimensional matrix and its variations. The first chapter focuses on the two-dimensional model and looks at two examples of the functional-divisional model: Time-Warner and Procter & Gamble. Chapters Two and Three describe variations on the matrix theme. Chapter Two focuses on the two-hat model, and Chapter Three focuses on the baton pass model. Chapters Four and Five complete Part One. Chapter Four describes the design issues that companies face when a matrix exists at several levels. This challenge is called the matrix within a matrix. Chapter Five describes some tools for balancing power between the two sides of the matrix. It includes responsibility charts for defining roles as well.

Part Two focuses on the more complex designs. We begin in Chapter Six with the business unit, geography, and functional three-dimensional design. Chapter Seven describes more complex forms, including the front-back model, which is replacing the three-dimensional global matrix in many situations. Chapter Eight describes IBM’s version of the front-back in some detail.

Part Three completes our discussion of the Star Model by presenting the planning processes, reward systems, and HR practices of successful implementers of the matrix. Chapter Nine focuses on the key communication practices. The planning and coordination processes are featured in Chapters Ten and Eleven. Chapter Twelve is devoted to the HR practices and rewards systems needed to reinforce the matrix behaviors. Leadership is key when implementing a matrix; Chapter Thirteen gives advice on the leader’s behaviors. Chapter Fourteen presents some approaches to managing the change process when implementing the matrix. In Chapter Fifteen I list the capabilities that companies need to implement a matrix successfully. They are mentioned individually throughout the book, but here I have gathered them together in one place as a summary. Matrix was one of those management practices that was initially hyped and then fell from grace. The Epilogue presents stories of what I have experienced in the use and abuse of the matrix.

Breckenridge, ColoradoAugust 2008Jay R. Galbraith

Introduction

MATRIX ORGANIZATIONS

What Are They? Where Did They Come From?

Why would a company choose a matrix type of organization structure and risk confusing people, such as those who have two bosses? It does so because its business strategies require it to be excellent simultaneously at two or three different things. For example, a business may need to create state-of-the-art technologies and then combine them rapidly into winning products. Strong functional or skill-specialty organizations foster new technologies and technical excellence. They attract talented people to the specialty and give them specialty careers. Yet competition requires the fast combination of specialties into new products and rapid time to market. The ability to meet these product development requirements is fostered by strong product or project managers. When company strategies require both technical excellence and fast time to market, a matrix of products and specialties results.

A multinational company may require excellence in global integration, local responsiveness, and functional skills. These companies require both global integration within a business and local responsiveness within a country. The global business unit coordinates across countries, and the country manager coordinates across businesses. The company does not want to choose between being global or being local. It wants to be global when launching new products and local when serving customers. Again, a matrix organization results when the company’s strategy requires both outcomes. The price of achieving both outcomes simultaneously is having to deal with the complexity of managing in a two-boss structure.

One of the main drivers for choosing a matrix structure is the pursuit of a dual- or multiple-priority strategy. The multinational company places equal priority on being global and on being local. It may pursue multiple priorities if it also requires functional excellence. A high-technology company like Nokia will want excellence in R&D and supply chain. Nokia values its brand and will want marketing excellence. These strategic intentions are implemented by having a talented person in charge of these functions. These functional people as well as global businesses and local geographies will report to the CEO or someone else high in the organization’s structure. A matrix structure results.

The other reason for choosing a matrix structure is the sharing of specialized and expensive resources. If the aforementioned business put all the specialists in the product organizations, it could achieve fast time to market. But specialists like to work in functional structures. And dedicating specialists to specific products is expensive. Companies usually want to avoid duplication and to be able to move people and ideas across products. The need to have strong product units that share functional specialists is another driver for adopting a matrix structure. Or the resource could be very expensive. It costs around $3 billion for a silicon fabrication facility today. Each business unit cannot afford its own. Instead the silicon fabs are in a separate unit and matrixed across the businesses.

I’ve written this book for two reasons. First, it is still a challenge to get a matrix organization to work smoothly. More and more managers are saying, “We have to learn how to make a matrix work.” These managers have learned that managing in a matrix is an essential capability that their company needs to learn. This book is intended to provide the guidance for building that capability. Second, there will be an increasing number of situations where matrix will be the organization of choice. All repetitive, well-understood types of work will be automated, go into software, or be outsourced to low-cost countries. The work left in our organizations will be professional work. It is this professional work that is appropriate for matrix types of organization. The bottom line is that we need to learn how to design matrix organizations that can function effectively.

In this Introduction, I first define what I mean by a matrix organization and give some examples. Then I briefly trace the origins of matrix up to the present time. Finally I present the Star Model, which will guide the complete design of the matrix type of organization. The rest of the book will examine the structures, processes, and HR practices that support a well-functioning matrix design.

What Is a Matrix?

A matrix is a type of organization structure that is built around two or more dimensions, such as functions, products, or regions, and in which people have two bosses. The structure shown in Figure I.1 is an example.

The figure shows two dimensions reporting to the R&D lab director. One dimension is the skill-specialty groups, such as electronics and software. The second dimension is projects that use the talent in the specialty groups in a changing array of R&D projects. The figure shows two projects for the automotive and the aerospace product lines. In each specialty group, there is a subproject manager who manages the project and the talent within the specialty group. The figure shows that there are two subproject managers in each specialty, one for auto and one for aero. This subproject manager reports to both the skill-specialty manager and to the project manager. This dual reporting is one of the defining characteristics of matrix organizations. It means that the subproject manager gets direction from both bosses. The dual reporting is also intended to imply a roughly equal power balance between the two sides of the matrix. This balance of power is the other defining characteristic of the matrix organization. Usually there is an attempt to carve out separate areas in which the two bosses will have authority. For example, the project manager will usually determine what to do and when to do it. The specialist manager will have jurisdiction over how to do the work. There is still overlap between the areas, but attempts to define areas of authority will clarify the situation and minimize disputes.

Figure I.1: A Matrix Structure of Specialists and Projects in an R & D Lab

When the matrix is working well, the two bosses communicate with each other to detect issues early and prevent unnecessary conflicts. Usually they jointly select the person to be the subproject manager. They then agree on a set of goals against which the manager will be evaluated. This agreement minimizes giving conflicting goals to the manager with two bosses. The subproject manager is then jointly evaluated, and both bosses sign off on the performance review. In this way, the subproject manager reports to two bosses.

A couple of other common examples are shown in Figures I.2 and I.3.

Figure I.2: Typical Corporate Structure with BusinessUnit CFO Reporting to the Business Unit and CFO

Figure I.3: Multibusiness, Multinational Matrix

In Figure I.2 is a matrix structure in a multibusiness corporation. Typically the functions (finance, HR, legal, and so on) report to the business unit president and to the corporate functional vice president. The finance function is shown in Figure I.2. Another common example is in international structures. Figure I.3 shows a multibusiness, multinational corporation’s structure. In this company, there are many businesses that are clustered into groups, sometimes called sectors. When the company is present in a large number of countries, the countries are grouped into regions. This structure then has two levels between the CEO and the manager with two bosses. In Figure I.2, there is only one level. The examples show the variations that are possible. In general, the fewer the layers between the two-boss manager and the leader who can resolve disputes, the better the matrix will function.

What Are the Origins of the Matrix?

In my view, the modern matrix emerged from the aerospace industry in the 1960s. However, it has its origins in the Scientific Management era of the early 1900s. It was Frederick Taylor who suggested the benefits of having multiple bosses. His idea was labeled functional foremanship. He suggested that the workforce have a schedule boss, a quality boss, a tool boss, an administrative boss, and so on. He wanted to bring specialist skills directly to the workforce. No one else liked this idea because of the confusion of multiple bosses.

The acceptable position was articulated in 1917 by Henri Fayol ([1949] 1987). He described the line-and-staff model. He underlined the concept of unity of command by stating that every employee should have one and only one boss. In addition, authority should run in an unbroken line from the leader at the top to the workers at the bottom. Together these ideas specify a hierarchy of authority as the preferred organizational form. This hierarchy was the line organization. However, he liked the idea of bringing expertise to bear through specialist roles. But these specialist roles, called staff roles, had no authority. They would provide advice and service when requested by the line managers in the hierarchy. When the line managers were persuaded by the staff, the line and not the staff would issue the expert advice as their command. Only the line could issue official orders. This model preserved the unity-of-command principle yet allowed expert advice to be used.

In practice, the line-and-staff model was never as clear as Fayol’s presentation of it. For example, the finance function shown in Figure I.2 would often act as if it were in command. When a company was financially stressed or when there was some accounting chicanery in the field, the communication from the CFO seemed less like advice and more like an order. However, the fiction of unity of command was always maintained. In the 1960s, that fiction was confronted in aerospace.

The aerospace industry grew rapidly after Russia launched Sputnik in 1957. A large number of projects were initiated. The Rand Corporation did a study of all these projects in the early 1960s (Peck and Scherer, 1962). It found that these projects met or exceeded their technical targets. Schedule performance was on average 150 percent in excess of the target completion date. Cost and budget performance was an average 320 percent above the initial estimates. These statistics clearly reflected the national priorities. Get the best technology into the field as quickly as possible and spend whatever it takes to meet the number one and two goals. It was a great time to be an aerospace engineer.

In the late 1960s, the national priorities changed. There were simply not enough resources to fund the three big national programs. First there was the space program. Schedules became a priority when Kennedy said “Man on the moon by 1970” (not 1975). Then there was the defense buildup to support the Vietnam War. At the same time, Boeing was growing to dominate the global commercial aircraft market. Following the introduction of the 707 and the 727, Boeing was launching the 737, the 747, and the government-supported Supersonic Transport (SST). When resources are scarce, costs and budgets become priorities. The result was that we still wanted state-of-the-art technology to get to the moon (and back). But aerospace companies had to deliver their projects more cost-effectively and on schedule. These were dual strategic priorities. NASA and the Department of Defense (DOD) changed from cost-plus to incentive contracts. The incentives were for cost and schedule performance outcomes as well as technical performance. There was a time when DOD and NASA even issued fixed-price contracts. It was a new game in aerospace. The change in strategic priorities resulted in changes in the aerospace company organizations. They all adopted matrix structures. They maintained the strong functions, such as engineering, but added a new strong project management organization. Before, projects were run by project coordinators who worked for the vice president of engineering. That structure represented the old priorities. The new dual priorities meant that strong project managers reported to the general manager along with the head of engineering. The structure created a power balance. The project managers became the champions of cost and schedule performance as well as technical performance on their projects. Although making the change was difficult, it was regarded as successful, and it contributed to the success of the NASA lunar landing program. There was a man on the moon before 1970.

The commercial world was initially reluctant to adopt the matrix. It was seen as expensive, and only the government could afford the overhead of the dual structure. But Boeing applied the matrix to its commercial airplane program, which is where I learned matrix from 1967 to 1969 (see Galbraith, 1973, chap. 9). The R&D community also began to adopt it in their labs. (Today matrix is seen as the natural way to run an R&D operation.) From the R&D experience it spread into the rest of the organization, especially for companies competing on new product development. The press then picked matrix up as the hot new idea. Books began to appear. The best was Stan Davis and Paul Lawrence’s Matrix (1977).

By the late 1970s, however, enough companies were experiencing difficulties that the word spread that “matrix doesn’t work.” The last nail in the coffin came from Peters and Waterman’s In Search of Excellence (1982). They claimed that none of the excellent companies used a matrix. The assertion was not true. Intel, Texas Instruments, Digital Equipment, Boeing, Fluor, and Bechtel all used it appropriately in their businesses. But that didn’t matter; the coffin was sealed. It would not be until the 1990s that managers could again talk openly in public about using a matrix organization. What happened?

What Happened?

Matrix organization is one of those management concepts, like Total Quality Management (TQM) or reengineering, that became very popular and then went through the management fashion cycle. Coming out of aerospace in the 1960s, matrix became popular in the 1970s and early 1980s. And, as usually happens, matrix in many of these organizations was wrongly adopted, hastily installed, and inappropriately implemented. When performance did not meet expectations, the matrix structure was abandoned. The word quickly spread: “We tried matrix, and it doesn’t work.” From then on, managers avoided matrix, even in situations where it was appropriate. Then they noticed two things. First, matrix seems to have worked at some companies. Second, they recognized that they faced situations where a matrix would be valuable. At that time, around the mid-to late 1990s, the concept of matrix assumed its normal place in management’s toolkit. That is where matrix is today, almost.

Matrix is a collaborative organizational form. I believe it needs to be implemented using a collaborative change process. The people should develop the collaborative skills during implementation that they will need during the full-scale operation. Instead, in the past many matrix structures were introduced through the command-and-control process. People were ordered to collaborate and usually faked it. The parts of the organization that had power were ordered to share it. They usually did not and actively (but usually passively) resisted the introduction of power-sharing arrangements. Then when the outcomes that required power sharing were not achieved, the resistors were quick to point out the failures and to suggest that the matrix organization was not working. So, quite often, poorly managed or unmanaged change processes resulted in failure, even when matrix was appropriate for the business situation.

Very often the company, division, or country did not have the capabilities to support a full-blown matrix structure. For example, if a division was characterized by functional silos, a matrix using product managers would not work. But recommendations to knock down the silos first were not appreciated. The division manager had two years to make his mark. He usually went ahead with a matrix structure anyway, to his chagrin. Thus many matrix organizations were hastily installed rather than carefully built over time.

Matrix was often inappropriately implemented. Once a new management concept becomes “acclaimed,” it becomes a tool in company power politics. For example, there were a large number of conglomerates in the 1970s. Except for the finance function, the corporate functions in a conglomerate were “lite” versions of corporate staffs. They were like Fayol’s staff roles. They were to deliver advice and service. If a business unit did not request service, the corporate functions were to leave it alone. In my opinion, as well as that of others, the lite form of corporate function is the appropriate form in a conglomerate structure. Many conglomerate corporate functions did not see it that way. Instead they invoked the matrix organization as the ideal modern structure. They would say, “We will never be among Fortune’s most admired companies unless we adopt the matrix form of organization.” Of course the real agenda was a power grab. The functions wanted a dotted or preferably a solid line into their functions within the business units. They could then make mischief in the businesses. Matrix was an ideal tool for these political maneuvers because it dealt directly with the power dimension. But such abuses led to more resistance to the concept of matrix organizations. Indeed matrix was often a politically correct reason for failure and discontinuance of the strong function program.

The other way that matrix organizations were inappropriately installed was that they were incompletely implemented. Most matrix organizations implemented only structural changes. By the time the leaders negotiated the dotted and solid lines, they had little appetite for making changes to the budgeting process or the compensation system. As it turns out, these supporting changes to the process infrastructure are more important to making a matrix work than the dotted-line structure. So when a collaborative matrix structure conflicted with a reward system that promoted individual rather than group efforts, the reward system won every time.

A number of multinational companies initially built their businesses country by country. Then when they introduced global business units, there were no accounting systems to track the performance of a business across countries. Each country had different IT systems, different inventory valuation systems, different depreciation schedules, and so on. There was no way to coordinate these global business units. The power rested with the countries a little longer. So a matrix requires not just dual reporting relations but dual accounting systems to support both sides of the matrix.

The lesson is that changing to a matrix structure requires complementary and reinforcing changes to the IT system, planning and budgeting processes, the performance management system, the bonus awards, the selection and development criteria, and so on. It requires a complete organization redesign. From these lessons, I devised the Star Model (Galbraith, 1977), which I present in the next section. The model is my guide to building organizational capabilities carefully and completely so as to implement a matrix successfully when it is appropriate. The Star Model can be used for any redesign. We will apply it in subsequent chapters to the design of matrix organizations.

The Star Model

The Star Model framework for organization design is the foundation on which a company bases its design choices. The framework consists of a series of design policies that are controllable by management and can influence employee behavior. The policies are the tools with which management must become skilled in order to shape the decisions and behaviors of their organizations effectively.

Figure I.4: The Star Model

In the Star Model, as shown in Figure I.4, design policies fall into five categories. The first is strategy, which determines direction. The second is structure, which determines the location of decision-making power. The third is processes, which have to do with the flow of information; they are the means of responding to information technologies. The fourth is rewards and reward systems, which influence the motivation of people to perform and to address organizational goals. The fifth category of the model is made up of policies relating to people (human resource policies), which influence and frequently define the employees’ mind-sets and skills.

Strategy

Strategy is the company’s formula for winning. The company’s strategy specifies the goals and objectives to be achieved as well as the values and missions to be pursued; it sets out the basic direction of the company. The strategy specifically delineates the products or services to be provided, the markets to be served, and the value to be offered to the customer. It also specifies sources of competitive advantage.