The Business Forecasting Deal E-Book

Michael Gilliland

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Serie: SAS Institute Inc
• Sprache: Englisch

Seitenzahl: 375

Veröffentlichungsjahr: 2010

Beschreibung

Practical-nontechnical-solutions to the problems of business forecasting Written in a nontechnical style, this book provides practical solutions to common business forecasting problems, showing you how to think about business forecasting in the context of uncertainty, randomness and process performance. * Addresses the philosophical foundations of forecasting * Raises awareness of fundamental issues usually overlooked in pursuit of the perfect forecast * Introduces a new way to think about business forecasting, focusing on process efficiency and the elimination of worst practices * Provides practical approaches for the non-statistical problems forecasters face * Illustrates Forecast Value Added (FVA) Analysis for identifying waste in the forecasting process Couched in the context of uncertainty, randomness, and process performance, this book offers new, innovative ideas for resolving your business forecasting problems.

Leseprobe

Table of Contents

Praise

WILEY & SAS BUSINESS SERIES

Title Page

Copyright Page

Dedication

Foreword

Foreword

Acknowledgements

Prologue

NOTES

CHAPTER 1 - Fundamental Issues in Business Forecasting

THE PROBLEM OF INDUCTION

THE REALITIES OF BUSINESS FORECASTING

THE CONTEST

WHAT IS DEMAND?

CONSTRAINED FORECAST

DEMAND VOLATILITY

INHERENT VOLATILITY AND ARTIFICIAL VOLATILITY

EVILS OF VOLATILITY

EVALUATING FORECAST PERFORMANCE

EMBARKING ON IMPROVEMENT

NOTES

CHAPTER 2 - Worst Practices in Business Forecasting: Part 1

WORST PRACTICES IN THE MECHANICS OF FORECASTING

MODEL “OVERFITTING” AND “PICK-BEST” SELECTION

CONFUSING MODEL FIT WITH FORECAST ACCURACY

ACCURACY EXPECTATIONS AND PERFORMANCE GOALS

FAILURE TO USE A NAÏVE MODEL OR ASSESS FORECAST VALUE ADDED

FORECASTING HIERARCHIES

OUTLIER HANDLING

NOTES

CHAPTER 3 - Worst Practices in Business Forecasting: Part 2

WORST PRACTICES IN THE PROCESS AND PRACTICES OF FORECASTING

POLITICS OF FORECASTING

BLAMING THE FORECAST

ADDING VARIATION TO DEMAND

EVANGELICAL FORECASTING

OVERINVESTING IN THE FORECASTING FUNCTION

FORECASTING PERFORMANCE MEASUREMENT AND REPORTING

FORECASTING SOFTWARE SELECTION

EDITORIAL COMMENT ON FORECASTING PRACTICES

NOTES

CHAPTER 4 - Forecast Value Added Analysis

WHAT IS FORECAST VALUE ADDED?

THE NAÏVE FORECAST

WHY IS FVA IMPORTANT?

FVA ANALYSIS: STEP-BY-STEP

FURTHER APPLICATION OF FVA ANALYSIS

CASE STUDIES

SUMMARY: THE LEAN APPROACH TO FORECASTING

NOTES

CHAPTER 5 - Forecasting without History

TYPICAL NEW PRODUCT FORECASTING SITUATIONS

NEW PRODUCT FORECASTING BY STRUCTURED ANALOGY

ORGANIZATIONAL REALIGNMENT

SUMMARY

NOTES

CHAPTER 6 - Alternative Approaches to the Problems of Business Forecasting

STATISTICAL APPROACH

COLLABORATIVE APPROACH

SUPPLY CHAIN ENGINEERING APPROACH

PRUNING APPROACH

SUMMARY

NOTES

CHAPTER 7 - Implementing a Forecasting Solution

WHY DO FORECASTING IMPLEMENTATIONS FAIL?

PREPROJECT ASSESSMENT

REQUESTING INFORMATION OR PROPOSALS

EVALUATING SOFTWARE VENDORS

WARNING SIGNS OF FAILURE

NOTES

CHAPTER 8 - Practical First Steps

STEP 1: RECOGNIZE THE VOLATILITY VERSUS ACCURACY RELATIONSHIP

STEP 2: DETERMINE INHERENT AND ARTIFICIAL VOLATILITY

STEP 3: UNDERSTAND WHAT ACCURACY IS REASONABLE TO EXPECT

STEP 4: USE FORECAST VALUE ADDED ANALYSIS TO ELIMINATE WASTED EFFORTS

STEP 5: UTILIZE MEANINGFUL PERFORMANCE METRICS AND REPORTING

STEP 6: ELIMINATE WORST PRACTICES

STEP 7: CONSULT FORECASTING RESOURCES

NOTES

CHAPTER 9 - What Management Must Know About Forecasting

APHORISM 1: FORECASTING IS A HUGE WASTE OF MANAGEMENT TIME

APHORISM 2: ACCURACY IS DETERMINED MORE BY THE NATURE OF THE BEHAVIOR BEING ...

APHORISM 3: ORGANIZATIONAL POLICIES AND POLITICS CAN HAVE A SIGNIFICANT IMPACT ...

APHORISM 4: YOU MAY NOT CONTROL THE ACCURACY ACHIEVED, BUT YOU CAN CONTROL THE ...

APHORISM 5: THE SUREST WAY TO GET A BETTER FORECAST IS TO MAKE THE DEMAND FORECASTABLE

APHORISM 6: MINIMIZE THE ORGANIZATION’S RELIANCE ON FORECASTING

APHORISM 7: BEFORE INVESTING IN A NEW SYSTEM OR PROCESS, PUT IT TO THE TEST

NOTES

Epilogue

Glossary

APPENDIX - Forecasting FAQs

Index

ADDITIONAL PRAISE FOR THE BUSINESS FORECASTING DEAL

“This book is a must read for demand planning practitioners and for organizational leaders and management dealing with demand forecasting. The book lays out the problem definition and issues dealing with forecasting, and provides a practical approach for implementing the demand planning process.”

—Shashi Tripathi, Vice President Product Management, AGNITY Healthcare

“We had a few ‘Aha!’ moments when we first came across Mike Gilliland’s work through webinars and conferences. In this book, Mike compiles and shares his insights of the realities of business forecasting and provides recommendations on how to navigate them. Some highlights:

• Managing forecast accuracy expectations.

• Streamlining existing forecasting processes using Forecast Value Added (FVA) methodology.

• Identifying erroneous “best practices” that contribute to flawed approaches.

The Business Forecasting Deal is well-written, easy to understand, and intuitively appealing not only to practitioners but their business partners as well. A must-read!”

—Kean Chew, Demand Planning Senior Manager & Brad Ragland, Demand Planning Team Leader, HAVI Global Solutions

“This book effectively addresses the frequent challenge of understanding and communicating the limitations of forecasting. I have learned to view a forecast as a risk management tool, but it can only be used to the fullest when we are able to leverage the forecast to make the best supply chain decisions. Mike Gilliland has identified ways for corporations to become more knowledgeable about their performance, be better equipped to manage expectations, and shown how to improve forecast results where opportunities exist. This is an exciting read for me as we are on the cusp of implementing a forecastability analysis, utilizing FVA to assess our performance. This book equips me to educate both internal users (our analysts), as well as external customers of the forecast on this new information.”

—Mark Hahn, Manager—Sales Forecasting and Analysis, Amway Corporation

“Truly exceptional in its simple and straightforward commentary on forecasting as practised in several organizations. With valuable insights on how to get the ‘bang for the buck’ in the intriguing world of business forecasting and suggestions to improve the quality of forecasts, the book is a must-read for all those involved in forecasting function and responsible for supply chain effectiveness in organizations.”

—Suren Palakkal, Senior Solution Consultant, MEB Consulting LLC, USA

“Insightful, entertaining, and a must for all planning professionals. Mike Gilliland is able to take complex concepts and theories and describe them for all to understand. This book should be read by planners and executives.”

—Mary Côté, Senior e-Business Consultant, DeltaWare Systems Inc.

WILEY & SAS BUSINESS SERIES

The Wiley & SAS Business Series presents books that help senior-level managers with their critical management decisions.

Titles in the Wiley and SAS Business Series include:

Activity-Based Management for Financial Institutions: Driving Bottom-Line Results by Brent Bahnub

Business Intelligence Competency Centers: A Team Approach to Maximizing Competitive Advantage by Gloria J. Miller, Dagmar Brautigam, and Stefanie Gerlach

Business Intelligence Success Factors: Tools for Aligning Your Business in the Global Economy by Olivia Parr Rud

Case Studies in Performance Management: A Guide from the Experts by Tony C. Adkins

CIO Best Practices: Enabling Strategic Value with Information Technology by Joe Stenzel

Credit Risk Assessment: The New Lending System for Borrowers, Lenders, and Investors by Clark Abrahams and Mingyuan Zhang

Credit Risk Scorecards: Developing and Implementing Intelligent Credit Scoring by Naeem Siddiqi

Customer Data Integration: Reaching a Single Version of the Truth by Jill Dyche and Evan Levy

Demand-Driven Forecasting: A Structured Approach to Forecasting by Charles Chase

Enterprise Risk Management: A Methodology for Achieving Strategic Objectives by Gregory Monahan

Fair Lending Compliance: Intelligence and Implications for Credit Risk Management by Clark R. Abrahams and Mingyuan Zhang

Information Revolution: Using the Information Evolution Model to Grow Your Business by Jim Davis, Gloria J. Miller, and Allan Russell

Marketing Automation: Practical Steps to More Effective Direct Marketing by Jeff LeSueur

Mastering Organizational Knowledge Flow: How to Make Knowledge Sharing Work by Frank Leistner

Performance Management: Finding the Missing Pieces (to Close the Intelligence Gap) by Gary Cokins

Performance Management: Integrating Strategy Execution, Methodologies, Risk, and Analytics by Gary Cokins

The Data Asset: How Smart Companies Govern Their Data for Business Success by Tony Fisher

The New Know: Innovation Powered by Analytics by Thornton May

Visual Six Sigma: Making Data Analysis Lean by Ian Cox, Marie A. Gaudard, Philip J. Ramsey, Mia L. Stephens, and Leo Wright

For more information on any of the above titles, please visit www.wiley.com.

Copyright © 2010 by SAS Institute, Inc. 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/permissions.

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 or for technical support, 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 books. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data

Gilliland, Michael.

The business forecasting deal : exposing the myths, eliminating bad practices, providing practical solutions / Michael Gilliland.

p. cm.

Includes bibliographical references and index.

ISBN 978-0-470-57443-0 (cloth)

1. Business forecasting. I. Title.

HD30.27.G55 2010

658.4’0355—dc22

2010007796

To my parents, Dennis and Joan Gilliland—much nicer people than I am.

Foreword

Tom Wallace

Many years ago, when I was young, I got promoted into a job that involved sales forecasting. When I told a friend of mine about it, his reply was, “Gee, that’s too bad. What did you do wrong?”

After a short while in the job, I began to understand what he meant. Forecasting is most often the ultimate no-win game. You’re almost always wrong (except on those rare, random occasions when actual sales come in exactly on forecast); you get beat up routinely for your “lousy forecasts”; and, unlike another nonfun activity—the annual budgeting process—you must go through the forecasting cycle at least every month, perhaps more often.

After more time in the job, I concluded that whoever called economics the dismal science never had a job doing sales forecasting. It was not a lot of fun. And things aren’t much different today, 30 years later.

Well, why not? After all, there’s a great deal of forecasting software now on the market and that should certainly help, right? And we’re better educated today than back then, with many more MBAs, MSs and PhDs available. Plus we’ve had all those years to learn from our mistakes. Things surely should have gotten better, right? It hasn’t happened; we haven’t learned much from our collective mistakes over time. A contributing factor is the unfortunate tendency to seek “the silver bullet”—to search for the best forecasting software in the belief that it will solve our forecasting problems.

Complexity and rate of change in most of our businesses has increased sharply over the years. It’s tougher out there today, and that’s not going to change.

Against that backdrop, Mike Gilliland’s new book is a breath of fresh air. It’s simple, straightforward, easy to read, insightful—and loaded with solid, practical advice on how to improve the forecasting process in your company. To me, the most valuable section is Chapter 4 on forecast value added (FVA). This is a method to evaluate (a) which parts of a forecasting process are adding value, (b) which elements are not cost effective, and (c) which parts of the process are making the forecasts worse. Parts b and c should be eliminated.

Gilliland poses the question, “How would you like to get better forecasts for free?” He then shows how to do this and relates the experiences of users of the FVA method: Intel, AstraZeneca, Tempur-Pedic and others. FVA is the lean manufacturing mind-set applied to forecasting, and that’s great.

To those of you whose job primarily involves forecasting, I say simply: You owe it to yourselves to read this book closely, carefully, and cover-to-cover. To others of you—in marketing, sales, supply chain, general management, and so on—I say that you also should read it, perhaps not as intently or completely as the forecasters but being certain to hit the high spots. There are value-adds in every chapter.

My prediction is that this book will come to be regarded as one of the very best in the field of forecasting literature. It’s a potential game changer; it may move forecasting into a space that’s more productive, more respected, and—dare I say—more fun. That would be huge.

In conclusion, let me pay Mike the biggest compliment I can: I wish I had written this book. Thanks, Mike.

TOM WALLACE Sedona, Arizona

Foreword

Anne G. Robinson

We do not live in a boring world.

Companies do not operate the same, markets do not respond the same way, and all customers definitely do not want the same things. It’s this inexhaustible variability—and the careful blending of historical data, market trends and domain expertise it requires—that makes demand forecasting such an interesting art.

But forecasting is not for the faint of heart. If you choose this profession prepare yourself for continual disappointment. Reality will never match your prediction—you will always find yourself ahead of—or behind—the true curve.

Fortunately, forecasting experts like Mike Gilliland generously share their expertise, allowing us quantitative analysts to narrow that gap and deliver better-than-average forecasts. Through his many contributions in the forecasting community, Mike has introduced the concept of forecast value added (FVA), a metric that enables us to measure the relative value of a forecast.

I work for a prominent high tech company. For several years, my team and I were responsible for creating and delivering the statistical forecast for the supply chain. Challenged by complexities due to a wide breadth of products, varying customer lead times, and high volatility in demand, we went beyond the boundaries of our statistical software to create analytical models more relevant for our industry. Additionally, leveraging best practices in consensus forecasting, we partnered with our marketing and planning counterparts to include domain expertise and market trends and collectively deliver a final forecast.

The accuracies of our forecasts across product lines, however, varied greatly. Given the complex nature of our business, this was expected. The true value of the forecasting process came from its relative contribution against doing nothing. Based on Mike’s experiences, we successfully introduced the forecast value added metric (along with forecast accuracy and bias) at the executive level. The FVA gave management a way to recognize and appreciate the value that was being delivered as a result of our forecasting process.

Throughout this book, Mike Gilliland explains the FVA method as well as many other tips and tricks to creating a competent forecasting capability—and avoiding some of the potential disasters along the way.

Enjoy the journey! I’m certainly glad the world isn’t boring.

ANNE G. ROBINSON, PHD Sr. Manager, Information and Data Strategy Customer Value Chain Management, Cisco

Acknowledgments

I started this book during a mid-career “test retirement” in early 2004, while trying to unwind and decompress from the labors and indignity of 18 years at publicly held American corporations. In April of 2004 the book was put on hold, however, when I accepted an offer to join SAS Institute, the world’s largest privately held software company. I had been a long time user of SAS software (since 1985) and a huge fan, having built forecasting, planning, and reporting systems in SAS for Oscar Mayer Foods as an Operations Research Analyst fresh out of graduate school. I had no idea what I was getting myself into—but in a good way.

In January 2010 SAS was named #1 on the Fortune magazine list of “100 Best Companies to Work For”—an honor much deserved. SAS founders Jim Goodnight and John Sall have nurtured an environment where employees can be focused on one thing—helping customers solve their business problems—and not wasting time catering to the misguided whims of Wall Street. My everlasting gratitude goes to Goodnight and Sall for creating this environment, and to Shiva Kommareddi, Director of Solutions Product Management, for hiring me on.

My experience at SAS has made this book much better than it would have been before. I’ve had the opportunity to work on forecasting problems with an array of exceptional SAS colleagues: Michael Leonard, Udo Sglavo, Jim Ferris, Charlie Chase, Chip Wells, Snurre Jensen, Jack Hymanson, Ed Katz, Rob Stevens, Phil Weiss, Pete Dillman, Andy Waclawski, Sam Guseman, Allan Manning, Bob Lucas, Terry Woodfield, Paddy Fahey, Mark Little, Tonya Balan, Mary Grace Crissey, Rajesh Selukar, Michele Trovero, Jerzy Brzezicki, Mahesh Joshi, Bob Davis, Evan Stubbs, Robin Way, Gul Ege, Tammi Kay George, Brenda Wolfe (now at ESRI), and others. Particular thanks to Charlie for his review of the manuscript, and to Tammi Kay for her boundless creativity in promoting the book.

Outside of SAS, I’ve had the good fortune of knowing and learning from a wide network of industry professionals, many of them met through association with the Institute of Business Forecasting (IBF), the International Institute of Forecasters (IIF), and APICS. My thanks to IBF’s Anish Jain and Constance Korol for providing numerous speaking opportunities at their conferences and webinars, and to Dr. Chaman Jain, IBF founder and editor of The Journal of Business Forecasting, for his manuscript comments and allowing use of my previously published work.

Many other friends and professional associates were able to review the manuscript, or have applied (and expanded on) the method of Forecast Value Added (FVA) Analysis at their own companies. These include Debbie Blackburn and Robert Bloomer of BB&T, Emily Rodriguez of Intel, Jack Harwell of RadioShack, Mark Hahn of Amway, Mary Côté of DeltaWare, Scott Roy of Wells Dairy, Dave Wehling of Toro, Jonathon Karelse of Yokohama Tire Canada, Eric Wilson of Tempur-Pedic, Curtis Brewer of Bayer CropScience, Shashi Tripathi of AGNITY Healthcare, Kalyan Sengupta of Chevron, Kean Chew and Brad Ragland of HAVI Global Solutions, Drew Prince of NCR, Suren Palakkal of MEB Consulting, Richard Herrin of Tredegar, Andrew Leu of USDL, Sharon M. Powell of RTI, Karen Miracle of Masonite, and Evelyn Jarrett. Thanks also to Len Tashman (Editor of Foresight: The International Journal of Applied Forecasting) and Jennifer Proctor (Editor of APICS Magazine) for providing access to previously published materials, and to Rob Miller of Covidien for use of his “Comet Chart.”

Several of my SAS colleagues have been involved in the production and marketing of the book, or otherwise have helped publicize my work to a broader audience. These include Julie Platt, Shelly Goodin, and Shelley Sessoms in Publications, Kristine Vick and Buffie Silva in Field Marketing, Faye Merrideth in Public Relations, Sara Smith in Analyst Relations, Alison Bolen and Diane Lennox in External Communications, and Blanche Phillips in Online Marketing. Jerry Oglesby, Larry LaRusso, and Carrie Vetter have given me speaking exposure at the annual F20xx Business Forecasting Conference held at SAS corporate headquarters every June. Sam Guseman, in addition to being a co-inventor of the structured analogy approach for new product forecasting, prepared the screen shots in Chapter 5.

Special thanks go to a number of individuals who have had particular influence on my forecasting career, or on this book:

• John LaBella, VP—Application Delivery at Gap, Inc. John gave me my first full-time forecasting job at Oscar Mayer/Kraft Foods in 1991. Many of the ideas in this book (particularly the use of “demand factors” in Chapter 5), evolved from the collaboration and visionary leadership he provided early in my career.

• Joe Mazel of Mazel Associates. Joe is a fabulous writer and editor who offered valuable comments on an early draft of this manuscript. Since meeting at an IBF conference in 2002, I’ve considered Joe my unofficial (and unpaid!) “publicist” for the many writing, speaking, and professional association connections he has provided me.

• Martin Joseph, Managing Director of Rivershill Consultancy, Ltd. I met Martin while serving together on the IBF Advisory Board, when he was Head of Information Management and Forecasting at AstraZeneca. We made an instant connection through our mutual interests in demand volatility and the application of statistical process control techniques in business forecasting.

• Meredith John, Product Manager at SAS. Meredith has been an invaluable ally in the development and marketing of SAS forecasting software since joining the company in 2007. Meredith provided a full review of the initial completed manuscript, and her feedback led to dramatic improvements in the focus and organization of the finished book. In addition, with her past publishing experience, Meredith helped me survive the annoyances of book design and production, and she gets credit for prototyping the layout of the cover.

• Stacey Hamilton, Acquisitions Editor at SAS Publications. Stacey guided me through the writing process and kept me (mostly) on plan. She also had to put up with my tantrums and appeals for “artistic integrity” any time there was a disagreement with the publisher. Stacey could definitely succeed as a hostage negotiator if this SAS gig ever falls through.

• Jessica Crews, Graphic Designer. Jessica brought to life the disturbing image in my head—that the practice of forecasting is not so far removed from snake-oil sales and circus sideshows. In addition to the cover art, Jessica has translated my other waking delusions into the colorful images that periodically appear in my blog, The Business Forecasting Deal (blogs.sas.com/forecasting).

• Anne Robinson, Sr. Manager—Information and Data Strategy at Cisco. Anne has championed the application of FVA analysis and helped expand my professional horizon through her leadership role at INFORMS (the Institute for Operations Research and the Management Sciences). Anne provided a valuable second foreword describing the use of FVA at Cisco.

• Tom Wallace. I was truly honored by Tom’s willingness to write a foreword for this book. He is a giant in the field of Sales & Operations Planning, and has made major contributions worldwide through his teaching and consulting, and his books with Bob Stahl. When the publisher asked me for Tom’s job title and company, I was aghast—that was like asking Cher or Madonna for their title and company! Most of us in the planning and forecasting fields would agree that Tom is just as much a celebrity as Cher or Madonna—even if he doesn’t have quite as fabulous a wardrobe.

• Anne Milley, Senior Director of Analytic Strategy at SAS. As my boss since late 2005, Anne has been a constant source of knowledge, inspiration, and positive feedback. She paved the way internally for me to write this book, and made many valuable comments on the developing manuscript. My best wishes to Anne as she embarks on her new role as the public voice of analytics at SAS.

The views expressed in this book are my own, and should not negatively reflect on the good taste and better judgment of the above mentioned reviewers, or of SAS Institute. All inaccuracies, hyperbolic claims, wild assertions, or outright lies, are my fault alone.

Lastly, I want to give very special thanks to Debbie Blackburn for, among many other things, six years of gentle nudging to resurrect this book project.

Prologue

Forecasting is a huge waste of management time.

The Business Forecasting Deal1 is written around this simple premise. It doesn’t mean that forecasting is pointless and irrelevant. It doesn’t mean that forecasting isn’t useful or necessary to run our organizations. It doesn’t mean that managers should not care about their forecasting issues, nor seek ways to improve them. It simply means that:

The amount of time, money, and human effort spent on forecasting is not commensurate with the amount of benefit achieved (the improvement in accuracy).

We spend far too many organizational resources creating our forecasts, while almost invariably failing to achieve the level of accuracy desired. The whole conversation needs to be turned around. We should be focusing much less on modeling and forecast accuracy and much more on process efficiency and effectiveness. We must also consider alternative solutions to the business problems that we, out of habit, rely on forecasting alone to address.

This book aims to expose the myths and bad practices in business forecasting, and to provide practical solutions. One such myth is that the desired level of forecast accuracy is always possible. The practicing forecaster soon realizes, however, that there are limits to the accuracy he or she can ever expect to achieve. This accuracy is largely determined by the forecastability of the behavior being forecast. Consider the simplest of examples—calling heads or tails in the toss of a fair coin. Over a large number of trials we cannot forecast the outcome (call the toss) correctly other than 50% of the time. Our accuracy has been determined by the nature of the behavior we are trying to forecast. As it is with coins, so it is (to a less obvious extent) with the things we attempt to forecast in business.

The reality is that smooth, stable, repeating patterns can be forecast accurately with simple techniques and little effort. Wild, volatile, and erratic patterns, however, may never be forecast accurately—no matter how elaborate the process and statistical sophistication we throw at the problem. In short:

We may never be able to control the accuracy achieved, or achieve the level of accuracy desired. But we can control the forecasting process we use, and the resources we invest.

This ties into a second myth, that the accuracy of our forecasts is proportional to the extent of our forecasting efforts. “If only,” management bemoans. “If only we had more data, a bigger computer, a more elaborate process, and better forecasters (or made the ones we have work harder!), we could get better forecasts.” But this is a false belief.

Curiously, there is often an inverse relationship between the amount of management attention given to forecasting and the accuracy of the results. The more a forecast is touched, the more it tends to go awry. Each process step, each opportunity to adjust a forecast, is just one more chance for wishes and politics and personal agendas to contaminate what should be an unbiased best guess at what is really going to happen. The purpose of this book is to explore why this happens—why there is such waste and inefficiency in the typical business forecasting process—and to suggest how to stop this from happening.

A third myth is that improving forecast accuracy is the ultimate goal—that improved accuracy is the best way, and perhaps even the only way, to improve organizational performance. But this belief can focus management’s attention on the wrong problem. Unless you work at a consulting firm selling forecasts:

The goal of your organization is not accurate forecasts—it is to make a profit and stay in business.

Forecast improvements are only a means to this end. Unfortunately, improvements may be impossible to deliver (when your demand is unforecastable), they may be too costly to implement (not worth the benefits), or they may even go unused—if management is unwilling to accept the reality of what a more accurate forecast is telling them. Focusing only on forecast improvements ignores other, nonforecasting, approaches that may more effectively solve the underlying business problem.

The Business Forecasting Deal aims to treat a gap in the literature by addressing the very foundations of business forecasting. Not in the careless and dogmatic way that we normally approach things, but critically, in a way to draw out all the subtlety and implication of our assumptions and beliefs.2 The book explores issues left unmentioned in the traditional forecasting literature—unmentioned because we assume to understand them, or don’t recognize them as issues at all.

For those seeking more advanced forecast modeling skills, there are plenty of good books covering the mathematics of forecasting, but these topics won’t be covered here. While advanced statistical and forecast modeling skills are useful in forecasting research and practice, these skills are neutralized when internal politics and personal agendas dominate an organization’s forecasting process. These skills are also neutralized when the behavior being forecast is essentially unforecastable. A big computer and fancy models aren’t going to help forecast the toss of a fair coin.

While you won’t need advanced modeling skills to get through this book, you should have a curiosity about why business forecasting is so maddening and maligned, along with a willingness to consider alternative solutions to the ensuing business problems. Forecasting is not an end in itself—it is one of many means to running an organization more efficiently and more profitably. In some situations forecasting is not the answer, or it is only an answer of last resort. We shouldn’t assume that there is always something we can do to improve a forecast. However, there are always things we can do to address the business problem—they may just not involve forecasting.

Although this book invokes a critical tone, it is not meant to discourage forecasting practitioners or to stifle innovation. Rather, it is a critique of the bad practices and snake-oil solutions proffered by many vendors of forecasting services and software. It is meant to expose forecasting’s myths and many pitfalls, so that the same mistakes don’t have to be repeated over and over again by each new practitioner. The forecasting profession does not need another cheerleader. But forecasting professionals, and those who rely on them, do need to be realistic about the limitations of what forecasting can ever be expected to achieve. There are no magic formulas and no miracle solutions, but the truth can be a lot harder to sell than the fiction.

In writing The Business Forecasting Deal, I’ve pieced together self-contained sections addressing specific business forecasting topics, along with practical solutions. Throughout I advocate use of a simple method called forecast value added (FVA) analysis. The FVA metric allows an organization to identify waste and inefficiency in its forecasting process, and has been adopted at a number of major corporations. Some of these adopters have spoken publicly of their findings, and their results are shared in a series of brief case studies in Chapter 4, Forecast Value Added Analysis. For those ready to apply FVA at their own organizations, this chapter provides step-by-step details on conducting the analysis.

In the narrow sense, the goal of forecasting is to generate better forecasts. But a more accurate forecast, by itself, has no value unless it is used to help an organization run more effectively. My objective is to do just this—to help organizations run more effectively. First, by offering an alternative framework for thinking about the problems of business forecasting. Second, by providing specific methods for addressing the challenges of business forecasting. And third, by encouraging readers to consider new ideas and creative approaches—but never to assume anything works until its effectiveness has been demonstrated.

This book, alone, will not make you the best forecaster you can be. But it will help you avoid becoming the worst forecaster you can be.

NOTES

1The Business Forecasting Deal is intended for both the forecasting practitioner and for management overseeing (or dealing with) the forecasting function. Sometimes terminology or formulas may appear that are not familiar to the reader, and are not thoroughly defined at the point they are used. Refer to the Glossary for explication of any of these unfamiliar concepts (which are shown in bold italics the first time they appear in the text).

2 For a major influence on this book’s approach to addressing the problems of business forecasting, see Bertrand Russell’s The Problems of Philosophy, Oxford University Press (1959). Russell delivers a concise and accessible introduction to methods of analysis that are too often ignored in the business world.

CHAPTER 1

Fundamental Issues in Business Forecasting

Forecasting is a difficult, thankless, and sometimes futile endeavor. When accuracy is not quite where everyone wants it to be, we react by making significant new investment in technology, process, and people to solve the business problem.1 While we live in an uncertain and largely unpredictable world, we prefer to operate under, as Makridakis and Taleb suggest, an “illusion of control.”2 We think a bigger computer, a fancier model, and a more elaborate process are all we need to get better forecasts, but the world doesn’t work that way.

THE PROBLEM OF INDUCTION

Scottish philosopher David Hume formed an early statement of the problem of forecasting in his 1748 book, An Enquiry Concerning Human Understanding. Hume was concerned with induction, the reasoning from particular facts to general conclusions. Forecasting is an example of this.

Hume observed that when we eat a piece of bread, it nourishes us. So we extrapolate our finite particular experiences to a general belief that bread nourishes. When we do this, we are in fact creating a forecast that the next piece of bread we eat will nourish us.

But Hume was a philosopher, so it was his job to be a critic and to question everything that we blindly believe. He was compelled to ask the question, What justification is there for this belief that bread nourishes? Just because bread has exhibited this nourishment characteristic in the past, is that any proof that bread will continue to nourish in the future? In the business forecasting world, just because customer X has always ordered product Y in some particular pattern, is that any proof that this behavior pattern will continue into the future? Of course not! Hume came to the skeptical conclusion that there is no proof that the future will behave like the past. He resolved that, in fact, we have no justification for our forecasts. Hume is correct. It has been over 250 years, and there is still no refutation of Hume’s argument, and no evidence that the business world is getting any easier to forecast.

We should not let Hume’s ultimate skepticism completely ruin our day, however. Just because we have no logical proof that our forecasts will be any good doesn’t mean we never ought to try. This historical perspective isn’t meant to completely discourage us. But it does provide a valuable reminder that we aren’t gods, we don’t have special powers of omniscience, and that we have no right to expect to consistently predict the future very well.

THE REALITIES OF BUSINESS FORECASTING

The unfortunate reality is that investment in the forecasting function is no guarantee of better results. There are often fundamental issues that impact an organization’s ability to forecast accurately, yet these are largely unrecognized or ignored. Until these issues are addressed, however, further investment in the forecasting function may be wasted. We begin by identifying several fundamental issues that must be dealt with in pursuit of our objective:

To generate forecasts as accurate and unbiased as anyone can reasonably expect them to be, and to do this as efficiently as possible.

So who is to blame for all the unrealistic expectations around forecast accuracy? Unfortunately, these unrealistic expectations are perpetuated by many in the profession (including those selling forecasting-related software or services) who know better, or who at least should know better. The dream of forecast accuracy is always easier to sell than the harsh reality.

The harsh reality is that predicting the future is a very difficult thing! As statisticians and forecast analysts, the best we can ever do is discover the underlying structure or rule guiding the behavior that is being forecast, to find a model that accurately represents the pattern of behavior, and then pray the behavior pattern doesn’t change in the future.

Assume we do discover the underlying structure of the behavior, we correctly model that structure in our forecasting software, and the structure does not change in the future. Should we then be able to achieve perfect forecasts? Unfortunately, the answer is no. In any complex business or social system (including things like the buying behavior of customers), there remains an element of randomness. Even though we know the underlying structure and model the behavior correctly, our forecast accuracy will still be limited by the amount of randomness, and no further improvement in accuracy will be possible. We can see how this works with a forecasting contest.

THE CONTEST

There are three processes to be forecast:

P10: The percentage of heads in the tossing of 10 fair coins

P100: The percentage of heads in the tossing of 100 fair coins

P1000: The percentage of heads in the tossing of 1000 fair coins.

Every day, the three processes will be executed: The coins will be tossed, and we have to predict the percentage of heads. What is our forecasted percentage of heads each day for each process? Can we forecast one process better than the others? What accuracy will we achieve? Are there any investments we can make (better software, bigger computer, more elaborate forecasting process, more skilled statistical analyst) to improve our accuracy?

Exhibit 1.1 One Hundred Trials of P10, P100, and P1000a

This isn’t meant to be a trick question, and it doesn’t take a doctorate in statistics to figure it out: The only rational forecast each day for each process is 50% heads. Exhibit 1.1 illustrates 100 daily trials of each of these processes. Since we are dealing with the independent tossing of fair coins, then, by definition, each process behaves according to the same underlying structure or rule—that over a large number of trials, each process will average about 50% heads. We fully understand the nature of each process, and we realize it makes no sense to forecast anything other than 50% heads each day for each process. However, as illustrated in the exhibit, the variation in the percentage of heads in each process is vastly different, as is the accuracy of our forecasts.

When there is a lot of randomness, or noise, in the behavior, we cannot expect to forecast it very accurately. Even when we know everything there is to know about the rules guiding the behavior, as we do here, the amount of randomness limits how accurate we can ever be. Also, in situations like these, any additional investment in the forecasting process would be a waste. There is nothing we could ever do to forecast P10 more accurately than P100, or P100 more accurately than P1000. The nature of each process, its underlying structure along with its random variability, determined the level of accuracy we were able to achieve.

Real life demand patterns are different from this in that the underlying mechanisms, knowable or unknowable, are not so simple as to be illustrated by independent tosses of a fair coin. Real life demand patterns may or may not have an underlying structure, we may or may not be able to discover and model that underlying structure, the underlying structure may or may not continue into the future, and there will be some degree of randomness.

What makes real life demand patterns so difficult to forecast is that the underlying mechanisms guiding their behavior may not be so apparent or may not even exist. Even if there is some structure to the historical pattern, it may not be obvious and can require good software or a skilled analyst to uncover it. But even then, even if we can discover and model the underlying behavior, there is no guarantee the behavior won’t change over time. As forecasters, why do we even bother to try?

The coin tossing contest illustrates that there are limits to the forecast accuracy we can achieve. We can’t assume that by applying more data, bigger computers, and more sophisticated software, or by exhorting our forecasters to work harder, we can always achieve the level of accuracy we desire. It is important to understand the limits of forecast accuracy, and to understand what level of accuracy is reasonable to expect for a given demand pattern. The danger is that if you do not know what accuracy is reasonable to expect, you can reward inferior performance, or you can waste resources pursuing unrealistic or impossible accuracy objectives. You can also miss opportunities for alternative (non-forecasting) solutions to your business problems.

WHAT IS DEMAND?

This book is about forecasting for products and services. It is not about forecasting the weather, or interest rates, or the outcome of sporting or political events. It is about forecasting the quantity of things people will buy or the quantity of services they will seek.

In business forecasting, we talk about demand every day. We don’t think much about our use of the word because it seems pretty straightforward. Demand is commonly characterized as, “what the customers want, and when they want it,” sometimes with the added proviso, “at a price they are willing to pay, along with any other products they want at that time.” So far, everything seems to make sense.

When we refer to demand, we usually mean unconstrained or true demand because we take no consideration of our ability to fulfill it. (Note: I will treat demand, unconstrained demand, and true demand as synonyms.) We use constrained demand to describe how much of true demand can be fulfilled (after incorporating any limitations on our ability to provide the product or service demanded). Thus, constrained demand ≤ demand.

A good forecast of demand, far enough into the future, allows an organization to invest in all and only the facilities, equipment, materials, and staffing that it needs to most profitably fulfill that demand. The value of a good demand forecast is readily apparent, and we valiantly load demand history into our software and statistical models to start the forecasting process. The common characterization of demand becomes problematic, however, once we try to operationalize it (that is, when we start to describe the specific, systematic way to measure it). We need an operational definition to provide true demand history to our forecasting models and to measure the accuracy of our unconstrained demand forecast. We need to know what true demand really is, but soon realize that it may be unobservable.

The nonchalant use of demand will not work. We know orders, we know shipments, and we know sales. We know calls handled at call centers, transactions processed at retail stores, and hours billed by consultants. We can track inventory, out of stocks, fill rates, backorders, and cancellations. We have all this data available to us, but none of it is the same as true demand. Consider the situation at a manufacturer:

Unfortunately, few organizations service their customers perfectly. As such, orders are not a perfect reflection of true demand. This is because when the order fill rate is less than 100%, orders are subject to all kinds of gamesmanship. Here are three examples:

1. An unfilled order may be rolled ahead (carried over) to a future time bucket.

2. If shortages are anticipated, customers may inflate their orders to capture a larger share of an allocation.

3. If shortages are anticipated, customers may withhold orders or direct their demand to alternative products or suppliers.

In the first example, demand (the rolled ahead order) appears in a time bucket later than when it was really wanted by the customer. Rolling unfilled orders causes demand to be overstated—the orders appear in the original time bucket and again in future buckets until the demand is filled or the order is cancelled.

In the second example, the savvy customer (or sales rep) has advanced knowledge that a product is scarce and will be allocated. If the allocation is based on some criterion, such as fill all orders at 50%, the customer simply orders twice its true demand and then hopes to receive what it really wanted in the first place.

The third example not only contaminates the use of orders to reflect true demand, but it can also cause significant financial harm to your business. In a period of chronic supply shortages (due to either supply problems or much higher than anticipated demand), customers may simply go elsewhere. Customers may truly want your product (so there is real demand), but it won’t be reflected in your historical data because no orders were placed. While orders are often perceived as equal to or greater than true demand, this third example shows that what is ordered may also be less than true demand.