Managing and Leading Software Projects E-Book

Richard E. Fairley

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 867

Veröffentlichungsjahr: 2011

Beschreibung

* The book is organized around basic principles of software project management: planning and estimating, measuring and controlling, leading and communicating, and managing risk. * Introduces software development methods, from traditional (hacking, requirements to code, and waterfall) to iterative (incremental build, evolutionary, agile, and spiral). * Illustrates and emphasizes tailoring the development process to each project, with a foundation in the fundamentals that are true for all development methods. * Topics such as the WBS, estimation, schedule networks, organizing the project team, and performance reporting are integrated, rather than being relegating to appendices. * Each chapter in the book includes an appendix that covers the relevant topics from CMMI-DEV-v1.2, IEEE/ISO Standards 12207, IEEE Standard 1058, and the PMI® Body of Knowledge. (PMI is a registered mark of Project Management Institute, Inc.)

Leseprobe

CONTENTS

Preface

1 Introduction

1.1 Introduction to Software Project Management

1.2 Objectives of This Chapter

1.3 Why Managing and Leading Software Projects Is Difficult

1.4 The Nature of Project Constraints

1.5 A Workflow Model for Managing Software Projects

1.6 Organizational Structures for Software Projects

1.7 Organizing the Project Team

1.8 Maintaining the Project Vision and the Product Vision

1.9 Frameworks, Standards, and Guidelines

1.10 Key Points of Chapter 1

1.11 Overview of the Text

Appendix 1A: Frameworks, Standards, and Guidelines for Managing Software Projects

1A.1 The CMMI-DEV-vl. 2 Process Framework

1A.2 ISO/IEC and IEEE/EIA Standards 12207

1A.3 IEEE/EIA Standard 1058

1A.4 The PMI Body of Knowledge

2 Process Models for Software Development

2.1 Introduction to Process Models

2.2 Objectives of This Chapter

2.3 A Development-Process Framework

2.4 Tailoring the System Engineering Framework for Software-Only Projects

2.5 Traditional Software Development Process Models

2.6 Iterative-Development Process Models

2.7 Designing an Iterative-Development Process

2.8 The Role of Prototyping in Software Development

2.9 Key Points of Chapter 2

Appendix 2A: Frameworks, Standards, and Guidelines for Software Development Process Models

2A.1 The CMMI-DEV-vl. 2 Technical Solution Process Area

2A.2 Development Processes in ISO/IEC and IEEE/EIA Standards 12207

2A.3 Technical Process Plans in IEEE/EIA Standard 1058

2A.4 The PMI Body of Knowledge

Appendix 2B: Considerations for Selecting an Iterative-Development Model

3 Establishing Project Foundations

3.1 Introduction to Project Foundations

3.2 Objectives of This Chapter

3.3 Software Acquisition

3.4 Requirements Engineering

3.5 Process Foundations

3.6 Key Points of Chapter 3

Appendix 3A: Frameworks, Standards, and Guidelines for Product Foundations

3A.1 The CMMI-DEV-vl.2 Process Areas for Requirements Development and Requirements Management

3A.2 Product Foundations in ISO/IEC and IEEE/EIA Standards 12207

3A.3 IEEE/EIA Standard 1058

3A.4 The PMI Body of Knowledge

4 Plans and Planning

4.1 Introduction to the Planning Process

4.2 Objectives of This Chapter

4.3 The Planning Process

4.4 The CMMI-DEV-vl.2 Process Area for Project Planning

4.5 A Minimal Project Plan

4.6 A Template for Software Project Management Plans, 130Techniques for Preparing a Project Plan

4.7 Key Points of Chapter 4

Appendix 4A: Frameworks, Standards, and Guidelines for Project Planning

4A.1 The CMMI-DEV-vl.2 Project Planning Process Area

4A.2 ISO/IEC and IEEE/EIA Standards 12207

4A.3 IEEE/EIA Standard 1058

4A.4 The PMI Body of Knowledge

Appendix 4B: Annotated Outline for Software Project Management Plans, Based on IEEE Standard 1058

4B.1 Purpose

4B.2 Evolution of Plans

4B.3 Overview

4B.4 Format of a Software Project Management Plan

4B.5 Structure and Content of the Plan

5 Project Planning Techniques

5.1 Introduction to Project Planning Techniques

5.2 Objectives of This Chapter

5.3 The Scope of Planning

5.4 Rolling-Wave Planning

5.5 Scenarios for Developing a Project Plan

5.6 Developing the Architecture Decomposition View and the Work Breakdown Structure

5.7 Guidelines for Designing Work Breakdown Structures

5.8 Developing the Project Schedule

5.9 Developing Resource Profiles

5.10 Resource-Gantt Charts

5.11 Estimating Project Effort, Cost, and Schedule

5.12 Key Points of Chapter 5

Appendix 5A: Frameworks, Standards, and Guidelines for Project Planning Techniques

A5.1 Specific Practices of the CMMI-DEV-vl.2 Project Planning Process Area

5A.2 ISO/IEC and IEEE/EIA Standards 12207

5A.3 IEEE/EIA Standard 1058

5A.4 The PMI Body of Knowledge

6 Estimation Techniques

6.1 Introduction to Estimation Techniques

6.2 Objectives of This Chapter

6.3 Fundamental Principles of Estimation

6.4 Designing to Project Constraints

6.5 Estimating Product Size

6.6 Pragmatic Estimation Techniques

6.7 Theory-Based Estimation Models

6.8 Regression-Based Estimation Models

6.9 Estimation Tools

6.10 Estimating Life Cycle Resources, Effort, and Cost

6.11 An Estimation Procedure

6.12 A Template for Recording Estimates

6.13 Key Points of Chapter 6

Appendix 6A: Frameworks, Standards, and Guidelines for Estimation

6A.1 Estimation Goals and Practices of the CMMI-DEV-vl.2 Project Planning Process Area

6A.2 ISO/IEC and IEEE/EIA Standards 12207

6A.3 IEEE/EIA Standard 1058

6A.4 The PMI Body of Knowledge

7 Measuring and Controlling Work Products

7.1 Introduction to Measuring and Controlling Work Products

7.2 Objectives of This Chapter

7.3 Why Measure?

7.4 What Should Be Measured?

7.5 Measures and Measurement

7.6 Measuring Product Attributes

7.7 Measuring and Analyzing Software Defects

7.8 Choosing Product Measures

7.9 Practical Software Measurement

7.10 Guidelines for Measuring and Controlling Work Products

7.11 Rolling-Wave Adjustments Based on Product Measures and Measurement

7.12 Key Points of Chapter 7

Appendix 7A: Frameworks, Standards, and Guidelines for Measuring and Controlling Work Products

7A.1 The CMMI-DEV-vl. 2 Monitoring and Control Process Area

7A.2 ISO/IEC and IEEE/EIA Standards 12207

7A.3 IEEE/EIA Standard 1058

7A.4 The PMI Body of Knowledge

7A.5 Practical Software and Systems Measurement (PSM)

Appendix 7B: Procedures and Forms for Software Inspections

7B.1 Conducting a Software Inspection

7B.2 The Defect Checklist

7B.3 Conducting an Inspection Meeting

8 Measuring and Controlling Work Processes

8.1 Introduction to Measuring and Controlling Work Processes

8.2 Objectives of This Chapter

8.3 Measuring and Analyzing Effort

8.4 Measuring and Analyzing Rework Effort

8.5 Tracking Effort, Schedule, and Cost; Estimating Future Status

8.6 Earned Value Reporting

8.7 Project Control Panel®

8.8 Key Points of Chapter 8

Appendix 8A: Frameworks, Standards, and Guidelines for Measuring and Controlling Work Processes

9 Managing Project Risk

9.1 Introduction to Managing Project Risk

9.2 Objectives of This Chapter

9.3 An Overview of Risk Management for Software Projects

9.4 Conventional Project Management Techniques

9.5 Risk Identification Techniques

9.6 Risk Analysis and Prioritization

9.7 Risk Mitigation Strategies

9.8 Top-A’ Risk Tracking and Risk Registers

9.9 Controlling the Risk Management Process

9.10 Crisis Management

9.11 Risk Management at the Organizational Level

9.12 Joint Risk Management

9.13 Key Points of Chapter 9

Appendix 9A: Frameworks, Standards, and Guidelines for Risk Management

9A.1 The CMMI-DEV-vl. 2 Risk Management Process Area

9A.2 ISO/EIC and IEEE/EIA Standards 12207

9A.3 IEEE/EIA Standard 1058

9A.4 The PMI Body of Knowledge

9A.5 IEEE Standard 1540

Appendix 9B: Software Risk Management Glossary

10 Teams, Teamwork, Motivation, Leadership, and Communication

10.1 Introduction

10.2 Objectives of This Chapter

10.3 Managing versus Leading

10.4 Teams and Teamwork

10.5 Maintaining Morale and Motivation

10.6 Can’t versus Won’t

10.7 Personality Styles

10.8 The Five-Layer Behavioral Model

10.9 Key Points of Chapter 10

Appendix 10A: Frameworks, Standards, and Guidelines for Teamwork and Leadership

10A.1 The CMMI-DEV-vl. 2 Framework Processes

10A.2 ISO/IEC and IEEE/EIA Standards 12207

10A.3 IEEE/EIA Standard 1058

10A.4 The PMI Body of Knowledge

10A.5 Other Sources of Information

11 Organizational Issues

11.1 Introduction to Organizational Issues

11.2 Objectives of This Chapter

11.3 The Influence of Corporate Culture

11.4 Assessing and Nurturing Intellectual Capital

11.5 Key Personnel Roles

11.6 Fifteen Guidelines for Organizing and Leading Software Engineering Teams

11.7 Key Points of Chapter 11

Appendix 11: Frameworks, Standards, and Guidelines for Organizational Issues,

A11.1 The CMMI-DEV-v1.2 Process Framework

A11.2 ISO and IEEE Standards 12207

A11.3 IEEE/EIA Standard 1058

A11.4 The PMI Body of Knowledge

Glossary of Terms

Guidance for Term Projects

Index

Press Operating Committee

Chair

Linda Shaferformer Director, Software Quality InstituteThe University of Texas at Austin

Editor-in-Chief

Alan ClementsProfessorUniversity of Teesside

Board Members

David Anderson,Principal Lecturer, University of PortsmouthMark J. Christensen,Independent ConsultantJames Conrad,Associate Professor, UNC CharlotteMichael G. Hinchey,Director, Software Engineering Laboratory, NASA Goddard Space Flight CenterPhillip Laplante,Associate Professor, Software Engineering, Penn State UniversityRichard Thayer,Professor Emeritus, California State University, SacramentoDonald F. Shafer,Chief Technology Officer, Athens Group, Inc.Evan Butterfield,Director of Products and ServicesKate Guillemette,Product Development Editor, CS Press

IEEE Computer Society Publications

The world-renowned IEEE Computer Society publishes, promotes, and distributes a wide variety of authoritative computer science and engineering texts. These books are available from most retail outlets. Visit the CS Store at http://computer.org/cspress for a list of products.

IEEE Computer Society / Wiley Partnership

The IEEE Computer Society and Wiley partnership allows the CS Press authored book program to produce a number of exciting new titles in areas of computer science, computing and networking with a special focus on software engineering. IEEE Computer Society members continue to receive a 15% discount on these titles when purchased through Wiley or at wiley.com/ieeecs

To submit questions about the program or send proposals please e-mail [e-mail] or write to Books, IEEE Computer Society, 10662 Los Vaqueros Circle, Los Alamitos, CA 90720-1314. Telephone [phone].Additional information regarding the Computer Society authored book program can also be accessed from our web site athttp://computer.org/cspress.

Copyright © 2009 by IEEE Computer Society. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey. Published simultaneously in Canada.

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, [phone], 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, [phone], fax [phone], or online at http://www.wiley.com/go/permission.

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.

For general information on our other products and services please contact our Customer Care Department within the United States at [phone], outside the United States at [phone] or fax [phone].

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data is available.

ISBN: 978-0-470-29455-0

PREFACE

Too often those who develop and modify software and those who manage software development are like trains traveling different routes to a common destination. The managers want to arrive at the customer ‘s station with an acceptable product, on schedule and within budget. The developers want to deliver to the users a trainload of features and quality attributes; they will delay the time of arrival to do so, if allowed. Sometimes the two trains appear to be on the same schedule, but often one surges ahead only to be sidetracked by traffic of higher priority while the other chugs onward. One or both may be unexpectedly rerouted, making it difficult to rendezvous en route and at the final destination.

Managers traveling on their train often wonder why programmers cannot just write the code that needs to be written, correctly and completely, and deliver it when it is needed. Software developers traveling on their train wonder what their managers do all day. This text provides the insights, methods, tools, and techniques needed to keep both trains moving in unison through their signals and switches and, better yet, shows how they can combine their engines and freight to form a single express train running on a pair of rails, one technical, the other managerial.

By reading this text and working through the exercises, students, software developers, project managers, and prospective managers will learn why

managing a large computer programming project is like managing any other large undertaking—in more ways than most programmers believe. But in many ways it is different—in more ways than most professional managers expect._

Readers will learn how software projects differ from other kinds of projects (i.e., construction, agricultural, manufacturing, administrative, and traditional engineering projects), and they will learn how the methods and techniques of project management must be modified and adapted for software projects.

Those who are, or will become managers of software projects, will acquire the methods, tools, and techniques needed to effectively manage software projects, both large and small. Software developers, both neophyte student and journeyman/jour- neywoman professional, will gain an increased understanding of what managers do, or should be doing all day and why managers ask them to do the things they ask/ demand. These readers will gain the knowledge they need to become project managers. Those students and software developers who have no desire to become project managers will benefit by gaining an increased understanding of what those other folks do all day and why the seemingly extraneous things they, the developers, are asked to do are important to the success of their projects.

This text is intended as a textbook for upper division undergraduates and graduate students as well as for software practitioners and current and prospective software project managers. Exercises are included in each chapter. Practical hints and guidelines are included throughout the text, thus making it suitable for industrial short courses and for self-study by practitioners and managers.

Chapters 1 through 3 provide the context for the remainder of the text: Chapter 1 provides an introduction to software project management; Chapter 2 covers process models for developing software-intensive systems; Chapter 3 is concerned with establishing the product foundations for software projects.

Chapters 4 through 10 cover the four primary activities of software project management:

Planning and estimating is covered in Chapters 4 through 6.Measuring and controlling is covered in Chapters 7 and 8.Managing risk is covered in Chapter 9.Leading, motivating, and communicating are covered in Chapter 10.

Chapter 11 covers organizational issues and concludes the text with a summary of 15 guidelines for organizing and leading software engineering teams.

For each topic covered, the approach taken is to present the full scope of activities for the largest and most complex projects and to show how those activities can be tailored, adapted, and scaled to fit the needs of projects of various sizes and complexities.

Learning objectives are presented at the beginning of each chapter and each concludes with a summary of key points from the chapter. Occasional sidebars elaborate the material at hand. An appendix to each chapter relates the topics covered in that chapter to four leading sources of information concerning management of software projects:

1. CMMI-DEV-v1.2 process framework

2. ISO/IEC and IEEE/EIA Standards 12207

3. IEEE/EIA Standard 1058

4. PMI’s Body of Knowledge (PMBOK®)

The text is consistent with the guidelines contained in PMBOK and ACM/IEEE curriculum recommendations.

Presentation slides, document templates, and other supporting material for the text and for term projects are available at the following URL: computer.org/book_extras/fairley_software_projects

Terms used throughout this text are defined in the Glossary at the end of the text. Topics, schedule, and a template for term projects follow the Glossary and included are some hypothetical projects that can be used as the basis for term projects in a course or as examples that practitioners and managers can use to gain experience in preparing software project management plans. Schedule and templates for deliverables for the hypothetic projects are also provided; electronic copies of templates and some software tools are provided at the URL previously cited. Alternatively, practitioners and managers can apply the templates and tools to a past, present, or future project.

A continued example for planning and conducting a project to build the software element of an automated teller system is presented to motivate and explain the material contained in each chapter.

As is well known, one learns best by doing. I strongly recommended that the exercises at the end of each chapter be completed and that progress through the material be accompanied by an extended exercise (i.e., a term project) to develop some elements a project plan for a real or hypothetical software project. The planning exercise can be based on an actual project that the reader has been, is currently, or will be involve in; or it can be based on one of the hypotheticals at the end of the text; or it can be based on a project assigned by the instructor. A week-by-week schedule for completing the term project on a quarter or semester basis is provided. Completion of the planning exercise will result in a report that contains elements similar to those presented in IEEE/EIA Standard 1058 for software project management plans.

The material can be presented in reading/lecture/discussion format or by assigned readings followed by classroom or on-l ine discussions based on the exercises and the term project.

I am indebted to the pioneers who surveyed the terrain, prepared the roadbed, laid down the tracks, and drove the golden spike so that our project trains can proceed to their destinations. Those pioneers include Fred Brooks, the intellectual father of us all; Winston Royce, who showed us systematic approaches to software development and management of software projects; Barry Boehm, who was the first to address issues of software engineering economics, risk management, and so much more; Tom DeMarco, the master tactician of software development, project management, and peopleware; and the many others who prepared the way for this text. I accept responsibility for any misinterpretations or misstatements of their work. My apologies to those I have failed to credit in the text, either through ignorance or oversight.

Thanks to Mary Jane Fairley, Linda Shafer, and the other reviewers of the manuscript for taking the time to read it and for the many insightful comments they offered. Special thanks to the many students to whom I have presented this material and from whom I have learned as much as they have learned from me.

Teller County, Colorado

RICHARD E.(DICK) FAIRLEY

1INTRODUCTION

In many ways, managing a large computer programming project is like managing any other large undertaking—in more ways than most programmers believe. But in many other ways it is different—in more ways than most professional managers expect.1

—Fred Brooks

1.1 INTRODUCTION TO SOFTWARE PROJECT MANAGEMENT

When you become (or perhaps already are) the manager of a software project you will find that experience to be one of the most challenging and most rewarding endeavors of your career. You, as a project manager, will be (or are) responsible for (1) delivering an acceptable product, (2) on the specified delivery date, and (3) within the constraints of the specified budget, resources, and technology. In return you will have, or should have, authority to use the resources available to you in the ways you think best to achieve the project objectives within the constraints of acceptable product, delivery date, and budget, resources, and technology.

Unfortunately, software projects have the (often deserved) reputation of costing more than estimated, taking longer than planned, and delivering less in quantity and quality of product than expected or required. Avoiding this stereotypical situation is the challenge of managing and leading software projects.

There are four fundamental activities that you must accomplish if you are to be a successful project manager:

1. planning and estimating,

2. measuring and controlling,

3. communicating, coordinating, and leading, and

4. managing risk.

These are the major themes of this text.

1.2 OBJECTIVES OF THIS CHAPTER

After reading this chapter and completing the exercises, you should understand:

why managing and leading software projects is difficult,the nature of project constraints,a workflow model for software projects,the work products of software projects,the organizational context of software projects,organizing a software development team,maintaining the project vision and product goals, andthe nature of process frameworks, software engineering standards, and process guidelines.

Appendix 1A to this chapter provides an introduction to elements of the following frameworks, standards, and guidelines that are concerned with managing software projects: the SEI Capability Maturity Model® Integration CMMI-DEV-v1.2, ISO/ IEC and IEEE/EIA Standards 12207, IEEE/EIA Standard 1058, and the Project Management Body of Knowledge (PMBOK®). Terms used in this chapter and throughout this text are defined in a glossary at the end of the text. Presentation slides for this chapter and other supporting material are available at the URL listed in the Preface.

1.3 WHY MANAGING AND LEADING SOFTWARE PROJECTS IS DIFFICULT

A project is a group of coordinated activities conducted within a specific time frame for the purpose of achieving specified objectives. Some projects are personal in nature, for example, building a dog house or painting a bedroom. Other projects are conducted by organizations. The focus of this text is on projects conducted within software organizations. In a general sense, all organizational projects are similar:

objectives must be specified,a schedule of activities must be planned,resources allocated,responsibilities assigned,work activities coordinated,progress monitored,communication maintained,risk factors identified and confronted, andcorrective actions applied as necessary.

In a specific sense, the methods, tools, and techniques used to manage a project depend on the nature of the work to be accomplished and the work products to be produced. Manufacturing projects are different from construction projects, which are different from agricultural projects, which are different from computer hardware projects, which are different from software engineering projects, and so on. Each kind of project, including software projects, adapts and tailors the general procedures of project management to accommodate the unique aspects of the development processes and the nature of the product to be developed.

Fred Brooks has famously observed that four essential properties of software differentiate it from other kinds of engineering artifacts and make software projects difficult2:

1. complexity,

2. conformity,

3. changeability, and

4. invisibility of software.

1.3.1 Software Complexity

Software is more complex, for the effort and the expense required to construct it, than most artifacts produced by human endeavor. Assuming it costs $50 (USD) per line of code to construct a one-million line program (specify, design, implement, verify, validate, and deliver it), the resulting cost will be $50,000,000. While this is a large sum of money, it is a small fraction of the cost of constructing a complex spacecraft, a skyscraper, or a naval aircraft carrier.

Brooks says, “Software entities are more complex for their size [emphasis added] than perhaps any other human construct, because no two parts are alike (at least above the statement level).”3 It is difficult to visualize the size of a software program because software has no physical attributes; however, if one were to print a one-million line program the stack of paper would be about 10 feet (roughly 3 meters) high if the program were printed 50 lines per page. The printout would occupy a volume of about 6.5 cubic feet. Biological entities such as human beings are of similar volume and they are far more complex than computer software, but there are few, if any, human-made artifacts of comparable size that are as complex as software.

The complexity of software arises from the large number of unique, interacting parts in a software system. The parts are unique because, for the most part, they are encapsulated as functions, subroutines, or objects and invoked as needed rather than being replicated. Software parts have several different kinds of interactions, including serial and concurrent invocations, state transitions, data couplings, and interfaces to databases and external systems. Depiction of a software entity often requires several different representations to portray the numerous static structures, dynamic couplings, and modes of interaction that exist in computer software.

A seemingly “small” change in requirements is one of the many ways that complexity of the product may affect management of a project. Complexity within the parts and in the connections among parts may result in a large amount of evolutionary rework for the “small” change in requirements, thus upsetting the ability to make progress according to plan. For this reason many experienced project managers say there are no small requirements changes. Size and complexity can also hide defects that may not be discovered immediately and thus require additional, unplanned corrective rework later.

1.3.2 Software Conformity

Conformity is the second issue cited by Brooks. Software must conform to exacting specifications in the representation of each part, in the interfaces to other internal parts, and in the connections to the environment in which it operates. A missing semicolon or other syntactic error can be detected by a compiler but a defect in the program logic, or a timing error caused by failure to conform to the requirements may be difficult to detect until encountered in operation. Unlike software, tolerance among the interfaces of physical entities is the foundation of manufacturing and construction; no two physical parts that are joined together have, or are required to have, exact matches. Eli Whitney (of cotton gin fame) realized in 1798 that if musket parts were manufactured to specified tolerances, interchangeability of similar (but not identical) parts could be achieved.

There are no corresponding tolerances in the interfaces among software entities or between software entities and their environments. Interfaces among software parts must agree exactly in numbers and types of parameters and kind of couplings. There are no interface specifications for software stating that a parameter can be “an integer plus or minus 2%.”

Lack of conformity can cause problems when an existing software component cannot be reused as planned because it does not conform to the needs of the product under development. Lack of conformity might not be discovered until late in a project, thus necessitating development and integration of an acceptable component to replace the one that cannot be reused. This requires unplanned allocation of resources and can delay product completion. Complexity may have made it difficult to determine that the reuse component lacked the necessary conformity until the components it would interact with were completed.

1.3.3 Software Changeability

Changeability is Brooks’s third factor that makes software projects difficult. Software coordinates the operation of physical components and provides the functionality in software-intensive systems.4 Because software is the most easily changed element (i.e., the most malleable) in a software-intensive system, it is the most frequently changed element, particularly in the late stages of a project. Changes may occur because customers change their minds; competing products change; mission objectives change; laws, regulations, and business practices change; underlying hardware and software technology changes (processors, operating systems, application packages); and/or the operating environment of the software changes. If an early version of the final product is installed in the operating environment, it will change that environment and result in new requirements that will require changes to the product. Simply stated, now that the new system enables me to do A and B, I would like for it to also allow me to do C, or to do C instead of B.

Each proposed change in product requirements must be accompanied by an analysis of the impact of the change on project work activities:

what work products will have to be changed?how much time and effort will be required?who is available to make the changes?how will the change affect your plans for schedule, budget, resources, technology, other product features, and the quality attributes of the product?

The goal of impact analysis is to determine whether a proposed change is “in scope” or “out of scope.” In-scope changes to a software product are changes that can be accomplished with little or no disruption to planned work activities. Acceptance of an out-of-scope change to the product requirements must be accompanied by corresponding adjustments to the budget, resources, and/or schedule; and/or modification or elimination of other product requirements. These actions can bring a proposed out-of-scope requirement change into revised scope.

A commonly occurring source of problems in managing software projects is an out-of-scope product change that is not accompanied by corresponding changes to the schedule, resources, budget, and/or technology. The problems thus created include burn-out of personnel from excessive overtime, and reduction in quality because tired people make more mistakes. In addition reviews, testing, and other quality control techniques are often reduced or eliminated because of inadequate time and resources to accomplish the change and maintain these other activities.

1.3.4 Software Invisibility

The fourth of Brooks’s factors is invisibility. Software is said to be invisible because it has no physical properties. While the effects of executing software on a digital computer are observable, software itself cannot be seen, tasted, smelled, touched, or heard. Our five human senses are incapable of directly sensing software; software is thus an intangible entity. Work products such as requirements specifications, design documents, source code, and object code are representations of software, but they are not the software. At the most elemental level, software resides in the magnetization and current flow in an enormous number of electronic elements within a digital device. Because software has no physical presence we use different representations, at different levels of abstraction, in an attempt to visualize the inherently invisible entity.

Because software cannot be directly observed as can, for example, a building under construction or an agricultural plot being prepared for planting, the techniques presented in this text can be used to determine the true state of progress of a software project. An unfortunate result of failing to use these techniques is that software products under development are often reported to be “almost complete” for long periods of time with no objective evidence to support or refute the claim; this is the well-known “90% complete syndrome” of software projects. Many software projects have been canceled after large investments of effort, time, and money because no one could objectively determine the status of the work products or provide a credible estimate of a completion date or the cost to complete the project. Sad but true, this will occur again. You do not want to be the manager of one of those projects.

1.3.5 Team-Oriented, Intellect-Intensive Work

In addition to the essential properties of software (complexity, conformity, changeability, and invisibility), one additional factor distinguishes software projects from other kinds of projects: software projects are team-oriented, intellect-intensive endeavors. In contrast, assembly-line manufacturing, construction of buildings and roads, planting of rice, and harvesting of fruit are labor-intensive activities; the work is arranged so that each person can perform a task with a high degree of autonomy and a small amount of interaction with others. Productivity increases linearly with the number of workers added; the work will proceed roughly twice as fast if the number of workers is doubled. Although labor- saving machines have increased productivity in some of these areas, the roles played by humans in these kinds of projects are predominantly labor-intensive.

Software is developed by teams of individuals who engage in creative problem solving. Teams are necessary because it would take too much time for one person to develop a modern software system and because it is unlikely that one individual would possess the necessary range of skills. Suppose, for example, that the total effort to develop a software product or system5 results in a productivity level of 1000 lines of code per staff-month (more on this later). A one million line program would require 1000 staff-months. Because effort (staff-months) is the product of people and time, it would require 1 person 1000 months (about 83 years) to complete the project.

A feasible combination of people and time for a 1000 staff-month project might be a team of 50 people working for 20 months but not 1000 people working for 1 month or even 200 people working for 5 months. The later proposals (1000 × 1 and 200 × 5) are not feasible because scheduling constraints among work activities dictate that some activities cannot begin before other work activities are completed: you can’t design (some part of a system) without some corresponding requirements, you should not write code without a design specification for (that part of) the system, you cannot review or test code until some code has been written, you cannot integrate software modules until they are available for integration, and so on.

Adding people to a software development team does not, as a rule, increase overall productivity in a linear manner because the increased overhead of communicating with and coordinating work activities among the added people decreases the productivity of the existing team. To cite Fred Brooks once again, the number of communication paths among n workers is n(n − 1)/2, which is the number of links in a fully connected graph. Five workers have 20 communication paths, 10 have 45 paths, and 20 have 190. Increasing the size of a programming team from 5 to 10 members might, for example, might increase the production rate of the team from 5000 lines of code per week to 7500 lines of code per week, but not 10,000 lines of code per week as would occur with linear scaling. In The Mythical Man-Month, Brooks described this phenomenon as Brooks’s law6:

Adding manpower to a late software project makes it later.

Brooks’s law is based on three factors:

1. the time required for existing team members to indoctrinate new team members,

2. the learning curve for the new members, and

3. the increased communication overhead that results from the new and existing members working together.

Brooks’s law would not be true if the work assigned to the new members did not invoke any of these three conditions.

A simile that illustrates the issues of team-oriented software development is that of a team of authors writing a book as a collaborative project; a team of authors is very much like a team of software developers. In the beginning, requirements analysis must be performed to determine the kind of book to be written and the constraints that apply to writing it. The number and skills of team members will constrain the kind and size of book that can be written by the available team of authors within a specified time frame. Constraints may include the number of people on the writing team, knowledge and skills of team members, the required completion date, and the word-processing hardware and software available to be used.

Next the structure of the book must be designed: the number of chapters, a brief synopsis of each, and the relationships (interfaces) among chapters must be specified. The book may be structured into sections that contain several chapters each (subsystems), or the text may be structured into multiple volumes (a system of systems). The dynamic structure of the text may flow linearly in time or it may move backward and forward in time between successive chapters; primary and secondary plot lines may be interleaved. An important constraint is to develop a design structure that will allow each team member to accomplish some work while other team members are accomplishing their work so that the work activities can proceed in parallel. Some books are cleverly structured to have multiple endings; readers choose the one they like.

Design details to be decided include the format of textual layout, fonts to be used, footnoting and referencing conventions, and stylistic guidelines (use of active and passive voice, use of dialects and idioms). Writing of the text occurs within a predetermined schedule of production that includes reviews by other team members (peer reviews) and independent reviews by copy editors (independent verification). Revisions determined by the reviews must be accomplished. The goal of the writing team is to produce a seamless text that appears to have been written by one person in a single setting.

A deviation from the planned narrative by one or more team members might produce a ripple effect that would require extensive revision of the text. If the completed book were software, a single punctuation or grammatical error in the text would render the book unreadable until the writers or their copy editor repaired the defect. An editor determines that each iteration of elements of the text satisfy the conditions placed on it by other elements (verification). Finally, reviews by critics and purchases by readers will determine the degree to which the book satisfies its intended purpose in its intended environment (validation).

The various development phases of writing (analysis, high-level design, detailed design, implementation, peer review, independent verification, revision, and validation) are creative activities and thus rarely occur in linear, sequential fashion. Conducting analysis, preparing and revising the design of the text, and production, review, and revision of the various parts may be overlapped, interleaved, and iterated. Team members must each do their assigned tasks, coordinate their work with other team members, and communicate ideas, problems, and changes on a continuous basis. The narrative above depicts a so-called Plan-driven approach to writing a book and, by analogy, to developing software. An alternative is to pursue an Agile approach by which the team members start with a basic concept and evolve the text in an iterative manner. This approach can be successful:

if the team is small, say five or six members (to limit the complexity of communication);if all members have in mind a common understanding of the desired structure of the text (i.e., a “design metaphor”);if there is a strict page limit and a completion date (the project constraints);if each iteration occurs in one or a few days (to facilitate ongoing revisions in structure; known as “refactoring”); andif a knowledgeable reader (known as the “customer”) is available to review each iteration and provide guidance for the contents of the next iteration.

In some cases, the team members may work in pairs (“pair programming”) to enhance synergy of effort.

In reality, most software projects incorporate elements of a plan-driven approach and an agile approach. When pursuing an agile approach, the team members must understand the nature of the desired product to be delivered, a design metaphor must be established, and the constraints on schedule, budget, resources, and technology that must be observed; thus some requirements definition, design, and project planning must be done. Those who pursue a plan-driven strategy often pursue an iterative (agile) approach to developing, verifying, and validating the product to be delivered; frequent demonstrations provide tangible evidence of progress and permit incorporation of changes in an incremental manner.

The approach taken in this text is to present a plan-driven strategy, based on iterative development, that is suitable for the largest and most complex projects, and to show how the techniques can be tailored and adapted to suit the needs of small, simple projects as well as large, complex ones. Process models for software development are presented in Chapter 2.

Over time humans have learned to conduct agricultural, construction, and manufacturing projects that employ teams of workers who accomplish their tasks efficiently and effectively.7 Because software is characterized by complexity, conformity, changeability, and invisibility, and because software projects are conducted by teams of individuals engaged in intellect-intensive teamwork, we humans are not always as adept at conducting software projects as we are at conducting traditional kinds of projects in agriculture, construction, and manufacturing. Nevertheless, the techniques presented in this text will help you manage software projects efficiently and effectively, that is, with economical use of time and resources to achieve desired outcomes.

Your role as project manager is to plan and coordinate the work activities of your project team so that the team can accomplish more working in a coordinated manner than could be accomplished by each individual working with total autonomy.

1.4 THE NATURE OF PROJECT CONSTRAINTS

Many of the problems you will encounter, or have encountered, in software projects are caused by difficulties of management and leadership (i.e., planning, estimating, measuring, controlling, communicating, coordinating, and managing risk) rather than technical issues (i.e., analysis, design, coding, and testing). These difficulties arise from multiple sources; some you can control as a project manager and some you can’t . Factors you can’t control are called constraints, which are limitations imposed by external agents on some or all of the operational domain, operational requirements, product requirements, project scope, budget, resources, completion date, and platform technology. Table 1.1 lists some typical constraints for software projects and provides brief explanations.

The operational domain is the environment in which the delivered software will be used. Operational domains include virtually every area of modern society, including health care, finance, transportation, communication, entertainment, business, and manufacturing environments. Understanding the operational domain in which the software will operate is essential to success. Operational requirements describe the

TABLE 1.1 Typical constraints on software projects

ConstraintExplanationOperational domainEnvironment of the usersOperational requirementsUsers’ needs and desiresProduct requirementsFunctional capabilities and quality attributesScientific knowledgeAlgorithms and data structuresProcess standardsWays of conducting work activitiesProject scopeWork activities to be accomplishedResourcesAssets available to conduct a projectBudgetMoney used to acquire resourcesCompletion dateDelivery date for work productsPlatform technologySoftware tools and hardware/software baseBusiness goalsProfit, stability, growthEthical considerationsServing best interests of humans and society

users’ view (i.e., the external view) of the system to be delivered. Some desired features, as specified in the operational requirements, may be beyond the current state of scientific knowledge, either at large or within your organization. Product requirements are the developers’ view (i.e., the internal view) of the system to be built; they include the functional capabilities and quality attributes the delivered product must possess in order to satisfy the operational requirements.

Process standards specify ways of conducting the work activities of software projects. Your organization may have standardized ways of conducting specific activities, such as planning and estimating projects, and measuring project factors such as conformance to the schedule, expenditure of resources, and measurement of quality attributes of the evolving product. In some cases the customer may specify standards and guidelines for conducting a project. Four of the most commonly used frameworks for process standards are the Capability Maturity Model Integration (CMMI), ISO/IEEE Standard 12207, IEEE Standard 1058, and the Project Management Body of Knowledge (PMBOK). Elements of these standards and guidelines are contained in appendixes to the chapters of this text.

The scope of a project is the set of activities that must be accomplished to deliver an acceptable product on schedule and within budget. Resources are the assets, both corporate and external, that can be applied to the project. Resources have both quality and quantity attributes; for example, you may have a sufficient number of software developers available (quantity of assets), but they may not have the necessary skills (quality of assets). The budget is the money available to acquire and use resources; the budget for your project may be constrained so that resources available within the organization cannot be utilized. The completion date is the day on which the product must be finished and ready for delivery. In some cases there may be multiple completion dates on which subsets of the final product must be delivered. The constrained delivery date(s) may be unrealistic.

Platform technology includes the set of methods, tools, and development environments used to produce or modify a software product. Examples include tools to develop and document requirements and designs, compilers and debuggers to generate and check the code, version control tools to track evolving versions of a project’s work products, and testing tools to aid in verify the software. Platform technology also includes the hardware processors and operating systems on which the software is developed and on which it will operate (which may be the same or different). One or more aspects of the platform technology may be obsolete or otherwise inappropriate for the work to be done.

Business goals may constrain your project to complete the product as soon as possible (to maximize short-term revenue), or to produce the highest possible quality (to maintain credibility with existing customers). Ethical considerations may constrain your project from delivering a product with known defects or from incorporating knowledge of a competitor’s product gained by unethical methods.

Some of the most difficult problems you will encounter in managing software projects arise from establishing and maintaining a balance among the constraints on project scope, budget, resources, technology, and the scheduled delivery date:

1. scope: the work to be done;

2. budget: the money to acquire resources;

3. resources: the assets to do the job;

4. technology: methods and tools to be used; and

5. delivery date: the date on which the system must be ready for delivery.

The initial balance among these factors is established in your initial project plan. The scope of your project may change during project execution because of changes to product requirements or other factors such as the budget or delivery date. The constraints on your budget, resources, and schedule may change because of internal factors in your organization, changes in the operational environment of the product to be delivered, or competitive pressures. Changes in project scope must always be accompanied by corresponding changes in schedule, budget, resources, and (perhaps) technology.

The constraints listed in Table 1.1 reduce the conceptual space available in which to plan and conduct your project. For example, it may not be possible to deliver a satisfactory product using 10 people for 12 months, but it might be possible if the schedule were extended to 15 months or if the number of people were increased from 10 to 15, or if the requirements for the product were reduced to the functionality that can be delivered with acceptable quality by 10 people in 12 months. In addition to the constraints listed in Table 1.1, there may be political and sociological factors that you cannot control.

Some of the first things you must do in managing a software project are:

1. establish the success criteria for your project,

2. clarify the constraints on the project and the product, and

3. determine whether there is a reasonable chance of meeting the success criteria within the constraints.

Constraints should be clarified to determine whether there is any flexibility or possibility of trade-offs among the constraints because fewer or looser constraints increase the options for planning and executing your project. There may be priorities among the success criteria of delivering an acceptable product on schedule and within budget; for example, delivering on schedule may be more important than the number of features delivered, or features delivered may be more important than cost. There may be additional success criteria, such as establishing a working relationship with a new customer, or developing a product architecture that provides a basis for developing future products, that is, developing a product-line architecture that consists of base elements and configurable elements.

Factors you will have (or should have) some influence over include:

1. establishing the success criteria,

2. negotiating the project constraints,

3. obtaining consensus among project stakeholders on an initial set of operational requirements, and

4. obtaining consensus among project stakeholders on an initial set of product requirements.

Factors you will have responsibility for include:

5. making initial estimates and plans;

6. maintaining a balance among requirements, schedule, and resources as the project evolves;

7. measuring and controlling the progress of the work;

8. leading the project team and coordinating their work activities;

9. communicating with stakeholders; and

10. managing risk factors that might interfere with, or prevent achieving a successful outcome.

The major activities of project management are planning and estimating, measuring and controlling, communicating and leading, and managing risk factors. Planning and estimating are concerned with determining the scope of activities that must be accomplished, estimating effort and schedule for the overall project, and developing estimates and plans for each major work activity. Planning for measurement involves establishing a data collection and reporting system that will be used to determine and report the actual status of work activities and work products on a continuing basis. Controlling involves applying corrective actions when actual status, as indicated by the measurements, does not agree with planned status.

Communicating involves establishing and maintaining adequate communication channels among all involved parties so that everyone is aware of progress and problems, and so that they are constantly reminded of the goals and success criteria for the project. Leading is concerned with providing direction to, removing roadblocks for, and maintaining the morale of project personnel.

Risk management is concerned with identifying risk factors (potential problems), both initially and on a continuing basis; monitoring identified risk factors; and engaging in risk mitigation activities such as preparing contingency plans and executing them when necessary.

1.5 A WORKFLOW MODEL FOR MANAGING SOFTWARE PROJECTS

The primary objective of a software project is to develop and deliver one or more acceptable work products within the constraints of required features, quality attributes, project scope, budget, resources, completion date, technology, and other factors. The work products to be delivered (e.g., object code, training materials, and installation instructions) result from the flow of intermediate work products that are produced by and flow through the work processes (requirements, design, source code, and test scenarios).

The model of project workflow used in this text is presented in Figure 1.1. All models, including the one in Figure 1.1, are abstractions of real situations that emphasize some aspects of interest and suppress details that are unimportant to the purposes of the model. Important details may be expressed in subordinate models. Subordinate models to Figure 1.1 are presented throughout this text.

Figure 1.1 indicates some of the processes that support the primary activity of Product Development; they include Verification and Validation (V&V), Quality Assurance of work processes and work products (QA), Configuration Management (CM), and others. Some supporting processes and their purposes are listed in Table 1.2. Each supporting process must be accomplished in accordance with a well-defined model for accomplishing the work activities of that process.

The model in Figure 1.1 is called a process model because it emphasizes work activities and the flow of work products among work activities. Each work activity in a process model produces one or more work products that provide inputs to subsequent work activities. By work product we mean any document produced by a software project (including the source code). Some work products are delivered to the customer (called deliverable work products), while others are intermediate work products developed to advance the creative problem-solving process in an orderly manner. Some of the work products of software projects are listed in Table 1.3.

FIGURE 1.1 A workflow model for managing software projects

TABLE 1.2Some supporting processes for software development

Supporting ProcessPurposeConfiguration managementChange control, baseline management, product audits, product buildsVerificationDetermining the degree to which work products satisfy the conditions placed on them by other work products and work processesValidationDetermining the degree of fitness of work products for their intended use in their intended environmentsQuality AssuranceDetermining conformance of work processes and work products to policies, plans, and proceduresDocumentationPreparation and updating of intermediate and deliverable work productsDeveloper trainingMaintaining adequate and appropriate skillsUser and operator trainingImparting skills needed to effectively use and operate systems

TABLE 1.3Some work-product documents produced by software projects

DocumentContent of DocumentProject planRoadmap for conducting the projectStatus reportsState of progress, cost, schedule, and qualityMemos and meeting minutesIssues, problems, recommendations, and resolutionse - Mail messagesOngoing communicationsOperational requirementsUser needs, desires, and expectationsTechnical specificationProduct features and quality attributesArchitectural design documentComponents and interfacesDetailed design specificationAlgorithms, data structures, and interface details of individual modulesSource codeProduct implementationTest planProduct verification criteria, test scenarios, and facilitiesReference manualProduct encyclopediaHelp messagesGuidance for usersRelease notesKnown issues, hints, and guidelinesInstallation instructionsGuidance for operatorsMaintenance guideGuidance for maintainers

As Michael Jackson has observed, the entire description of a software system or product is usually too complex for the entire description to be written directly in a programming language, so we must prepare different descriptions at different levels of abstraction, and for different purposes [Jack02]. Note that each of the work products listed in Table 1.3 is a document; software developers and software project managers do not produce physical artifacts other than documents, which may exist in printed or electronic form.

As illustrated in the workflow model depicted in Figure 1.1, a software project is initiated by customer and managers. A customer is the person or organization that provides the requirements for and accepts the deliverable work products. Customers may place constraints on a project, such as specifying a required database interface (a product constraint) or the date when the delivered system must be available for use (a process constraint). Managers include your management and you, the project manager. Managers specify constraints and directives. A process constraint from your manager might place a limit on the number of people available to conduct the project; a management directive might require that all software projects in the organization perform a design activity. You, the project manager, might issue directives requiring that the design be documented using UML (the Universal Modeling Language) and that one or more design reviews be held.

Requirements, constraints, and directives provide the inputs to the planning process, which is (or should be) a group activity led by you, the project manager. You should involve the customer, selected members of the development team, and other primary stakeholders in the planning process. Planning involves estimation. Factors to be initially estimated include a schedule for conducting the major work activities; kinds and numbers of resources needed, when they will be needed, and for how long; and the project milestones (points in time when progress is assessed). Estimation is best accomplished by using historical data from a data repository. Data at the completion of your project can be placed in a repository to aid in estimation of future projects. Intermediate data can be retained to assess progress and prepare completion estimates, which may result in replanning.

The output of your planning process will include identification of the roles to be played in conducting the project, which results in assignment of personnel to those roles. During initial planning, the major work activities to be planned include software development and the various supporting processes such as configuration management, process and product quality assurance, verification, validation, user training; plus other necessary activities that constitute the scope of your project. Detailed plans for these activities will evolve as the project evolves.

During execution of the project, data are collected and status reports are prepared on a periodic basis by you and your staff. The status reports will be used by you (the project manager), your customer, your managers, support groups, and other project stakeholders. Status reports compare planned progress to actual progress; they may cause you and your customer, working together, to revise plans and requirements, or you might, for example, reassign some personnel to different project roles (e.g., a software designer might be moved to the independent validation team). Status data are also used to provide a basis for estimating future progress based on progress to date (which may result in replanning), and is retained to provide a basis of estimation for future projects.

Problem reports are generated to document defects discovered in work products that must be reworked. Status reports, new requirements, and changes to requirements, constraints, directives, and problem reports provide the data needed to continually update, elaborate, and revise your project plan.

Every organization that develops and maintains software, including yours, should have one or more workflow models of software development that depicts the major work activities and flow of work products. Each member of the organization should be familiar with the workflow model(s) and understand the ways in which their work activities and work products fit into the model(s). Everyone in your software development organization should be able to sketch and describe the workflow model(s) used in the organization. If there is more than one workflow model, everyone should understand the kinds of projects for which the various models are appropriate.

1.6 ORGANIZATIONAL STRUCTURES FOR SOFTWARE PROJECTS

Projects are one-time, transient events that are initiated to accomplish a specific purpose and are terminated when the project objectives are achieved (and are sometimes cancelled before achieving the objectives). A project exists within the context of the organization in which it is conducted; each project must adhere to the structural model of the organization. Departments that conduct engineering projects, including software projects, are typically organized in one of four ways: functional structure, project structure, matrix structure, or hybrid structure.

1.6.1 Functional Structures

As the name implies, workers in a functional organization are grouped by the functions they perform. Functional groups can be process-oriented or product-oriented. One process-oriented functional group might, for example, specialize in requirements engineering, another in design of user interfaces, another in design and implementation of code, another in product validation, and yet another in user training. When organized by product specialty, one group might specialize in data communication, another in database systems, another in user interfaces, and yet another in numerical algorithms. Figure 1.2 illustrates a process-oriented functional organization, and Figure 1.3 illustrates a product-oriented functional group.

Each functional group has a functional manager whose job is to acquire and maintain the quantity and quality of workers needed to support the projects within the organization, train them as necessary, provide the necessary tools, and coordinate their work activities on various projects. Different group members apply their

FIGURE 1.2 A process-oriented functional organization

FIGURE 1.3 A product-oriented functional organization

expertise to different projects as needed. As a project manager in a functional organization, responsible for delivering an acceptable product on schedule and within budget, your ability to successfully conduct your project will depend on your skill in working with the functional managers and their team members to complete the various work activities and develop the various work products for your project.

1.6.2 Project Structures

In a purely project-structured organization, you, as project manager, have full authority and responsibility for managing budget and resources. You acquire the kinds of workers you need to conduct your project and all project members report directly to you; you might acquire your workers from functional groups or you might hire them from outside. You, the project manager, have the authority to acquire staff members within the constraints of your budget and to remove them when they are no longer needed or are not performing up to your expectations. Your ability to successfully conduct your project depends on acquiring the quantity and quality of workers needed, training them as necessary, providing the necessary tools, and coordinating their work activities. A project-structured organization is illustrated in Figure 1.4.

1.6.3 Matrix Structures

The goal of a matrix organization is to obtain the advantages of both functional and project structures; functional specialists are assigned to projects as needed and work for you, the project manager, while applying their expertise to your project. When their tasks are completed, they return to their function groups and are assigned, as needed, to other projects. Workers in a matrix organization thus have two bosses: their functional manager and their project manager.

An example of a matrix organization is illustrated in Figure 1.5. The functional groups might be, for example, a user interface group, an algorithms group, a database group, and a communications protocol group. The numbers in the matrix indicate the number of workers of each functional type assigned to each project; for example, project #1 has 10 members: 2 of functional type #1 (user interface), 5 of functional type #3 (database), and 2 of functional type #4 (communications). Project #3 is the largest; it has 23 members. Currently 6 members of the user interface group are assigned to this project, 8 from the algorithms group, 2 from the database group, and 7 from communications.

Matrix organizations can be characterized as weak or strong, depending on the relative authority of the functional managers and the project managers. In a strong

FIGURE 1.4 A project-oriented organization

FIGURE 1.5 A matrix-structured organization

matrix, the functional managers have authority to assign workers to projects, and project managers must accept the workers assigned to them. In a weak matrix, the project manager controls the project budget, can reject workers from functional groups and hire outside workers if functional groups do not have sufficient quantities or qualities of workers.

When a matrix organization performs as intended, functional workers apply their specialties to different projects, under the direction of project managers, over time while retaining membership in a group of like-minded experts. Two problems that can occur in matrix organizations are (1) conflicts between functional managers and project managers over the allocation of worker resources (which puts the workers in untenable situations), and (2) frequent shifting of workers from project to project as crises occur (know as “firefighting” mode).

1.6.4 Hybrid Structures