THE

BUSINESSFORECASTING

DEALTH

E BUSIN

ESS FOR

ECA

STING

DEA

L

Exposing Myths, Eliminating Bad Practices, Providing Practical Solutions

Michael Gilliland

Gillilan

d

$49.95 USA / $59.95 CAN

THE BUSINESS

FORECASTING DEAL

Exposing Myths, Eliminating Bad Practices, Providing Practical Solutions

Business forecasting is the ultimate no-win game. You’re almost always wrong—except on those rare, random occasions where actual sales come in exactly on forecast—and you get beat up routinely for your “lousy forecasts.” Against this framework, The Business Forecasting Deal turns forecasting on its head. Instead of focusing solely on statistical modeling and forecast accuracy—accuracy being largely determined by the nature of the demand you are trying to forecast—Michael Gilliland turns your attention to forecasting process effi ciency and effectiveness. His unique perspective, utilizing the emerging method of Forecast Value Added (FVA) analysis, shows how organizations can meaningfully improve their performance by eliminating the “worst practices” that now sabotage and confound their forecasting efforts.

While predicting the future is a very diffi cult thing, there is no shortage of articles, books, consultants, and even software vendors willing to tell you (or sell you) their version of forecasting best practices. However, many of these purported “best practices” do not work—they fail to make the forecast any better. Through FVA analysis—a method that has been employed at several major corporations including Intel, AstraZeneca, Cisco, and Yokohama Tire (Canada)—The Business Forecasting Deal shows how to identify the waste and ineffi ciencies in the typical forecasting process. By eliminating those (surprisingly common) practices that make the forecast worse, FVA analysis is helping companies get better forecasts with less effort and less cost.

Written in a nontechnical style, this book provides practical solutions to a wide variety of business forecasting problems. It illustrates a new way to think about business forecasting in the context of uncertainty, randomness, and process performance, all within the internal political arena in which real-life forecasting is conducted. While there is no magic formula to guarantee perfect forecasts, this book will change your perspective—and improve your success—in dealing with the problems of business forecasting.

MICHAEL GILLILAND is Product Marketing Manager at SAS Institute and has worked in consu-mer products forecasting for more than twenty years. Prior to joining SAS in 2004, Mike held forecasting management positions in the food, electronics, and apparel industries and served as a consultant. He is a frequent speaker at industry events, has published articles in Supply Chain Management Review, Journal of Business Forecasting, Foresight, and APICS magazine, and was a columnist on “Worst Practices in Business Forecasting” for Supply Chain Forecasting Digest. Mike holds a BA in philosophy from Michigan State University, and master’s degrees in philosophy and mathematical sciences from Johns Hopkins University. Follow his blog, The Business Forecasting Deal, at blogs.sas.com/forecasting.

– continued on back f lap –

– continued from front f lap –

Jacket Design: Michael RutkowskiCover Illustration: Jessica Crews

Praise for

THE BUSINESS FORECASTING DEALExposing My ths, El iminat ing Bad Pract ices , Prov id ing Pract ica l Solut ions

“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.” —From the Foreword by Tom Wallace

“Into a market awash in forecasting texts, each promising to be the panacea for supply chain ineffi ciencies and sales shortfalls, Mike Gilliland’s The Business Forecasting Deal is a refreshingly objective assessment of the real costs and benefi ts of forecast process improvement. Above all, his book cuts to what should be at the heart of all our efforts: are the expenditures in time and money we are making in forecast improvement truly resulting in incremental profi tability?” —Jonathon Karelse, National Marketing Manager, Yokohama Tire (Canada) Inc.

“A great book, unique as well. Mike Gilliland is the fi rst person with a mission to bring ‘worst practices’ into the open so that forecasters can see what they are doing wrong and the price they are paying for it. Nothing can happen unless they recognize their problems, for which Mike has made a tremendous contribution.” —Dr. Chaman Jain, St. John’s University; Editor, Journal of Business Forecasting

“Mike Gilliland has captured the key issues in this book, and his work, particularly on volatility and Forecast Value Added, is to be applauded.” —Martin Joseph, Managing Director, Rivershill Consulting Ltd.

“I have used Mike Gilliland’s Forecast Value Added (FVA) analysis for the past year, and it has made a big difference in understanding what steps in our forecasting process are worthwhile and what steps need to be improved or eliminated. His advice on the proper way to present FVA to upper management has helped gain the confi dence of the division in this analysis.” —Dave Wehling, Manager, Marketing Requirements, International Division, The Toro Company

“Mike Gilliland provides a unique perspective on the forecasting discipline. He presents sound quantitative lessons paired with pragmatic, realistic, and effective methodologies for applying his techniques. This book is an excellent and perhaps mandatory addition to the repertoire of demand managers and supply chain management professionals alike.” —Emily Rodriguez, Supply Chain Program Manager, Intel Corporation

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

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 defi nition and issues

dealing with forecasting, and provides a practical approach for imple-

menting the demand planning process.”

—Shashi Tripathi, Vice President Product Management,

AGNITY Healthcare

“We had a few ‘Aha!’ moments when we fi rst came across Mike

Gilliland’s work through webinars and conferences. In this book,

Mike compiles and shares his insights of the realities of business fore-

casting 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 fl awed

approaches.

The Business Forecasting Deal is well-written, easy to understand, and

intuitively appealing not only to practitioners but their business part-

ners 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 understand-

ing 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 identifi ed 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 busi-

ness 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.

The Business Forecasting Deal

WILEY & SAS BUSINESS SERIES The Wiley & SAS Business Series presents books that help senior - level manag-ers 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 .

The Business Forecasting Deal

Exposing Myths, Eliminating Bad Practices, Providing

Practical Solutions

Michael Gilliland

John Wiley & Sons, Inc.

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, (978) 750-8400, fax (978) 646-8600, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, 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 specifi cally disclaim any implied warranties of merchantability or fi tness 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 profi t 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 (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

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

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

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

Contents

Foreword—Tom Wallace xiii

Foreword—Anne G. Robinson xv

Acknowledgments xvii

Prologue 1

Chapter 1 Fundamental Issues in Business Forecasting ..........................................5

The Problem of Induction 5

The Realities of Business Forecasting 6

The Contest 7

What Is Demand? 10

Constrained Forecast 13

Demand Volatility 15

Inherent Volatility and Artifi cial Volatility 17

Evils of Volatility 19

Evaluating Forecast Performance 22

Embarking on Improvement 24

Notes 26

Chapter 2 Worst Practices in Business Forecasting: Part 1 .................................................................29

Worst Practices in the Mechanics of Forecasting 30

Model “Overfi tting” and “Pick-Best” Selection 32

Confusing Model Fit with Forecast Accuracy 41

Accuracy Expectations and Performance Goals 43

Failure to Use a Naïve Model or Assess Forecast Value Added 47

ix

x ▸  C O N T E N T S

Forecasting Hierarchies 48

Outlier Handling 50

Notes 54

Chapter 3 Worst Practices in Business Forecasting: Part 2 .................................................................55

Worst Practices in the Process and Practices of Forecasting 55

Politics of Forecasting 57

Blaming the Forecast 60

Adding Variation to Demand 61

Evangelical Forecasting 64

Overinvesting in the Forecasting Function 66

Forecasting Performance Measurement and Reporting 69

Forecasting Software Selection 74

Editorial Comment on Forecasting Practices 76

Notes 78

Chapter 4 Forecast Value Added Analysis.........................81

What Is Forecast Value Added? 82

The Naïve Forecast 83

Why Is FVA Important? 90

FVA Analysis: Step-by-Step 92

Further Application of FVA Analysis 101

Case Studies 102

Summary: The Lean Approach to Forecasting 107

Notes 108

Chapter 5 Forecasting without History ...........................111

Typical New Product Forecasting Situations 111

New Product Forecasting by Structured Analogy 114

Organizational Realignment 120

Summary 131

Notes 132

C O N T E N T S  ◂ xi

Chapter 6 Alternative Approaches to the Problems of Business Forecasting ..................................133

Statistical Approach 134

Collaborative Approach 136

Supply Chain Engineering Approach 142

Pruning Approach 145

Summary 149

Notes 150

Chapter 7 Implementing a Forecasting Solution ............151

Why Do Forecasting Implementations Fail? 151

Preproject Assessment 153

Requesting Information or Proposals 154

Evaluating Software Vendors 155

Warning Signs of Failure 157

Notes 159

Chapter 8 Practical First Steps ........................................161

Step 1: Recognize the Volatility versus Accuracy Relationship 161

Step 2: Determine Inherent and Artifi cial Volatility 165

Step 3: Understand What Accuracy Is Reasonable to Expect 166

Step 4: Use Forecast Value Added Analysis to Eliminate Wasted Efforts 167

Step 5: Utilize Meaningful Performance Metrics and Reporting 168

Step 6: Eliminate Worst Practices 169

Step 7: Consult Forecasting Resources 170

Notes 173

Chapter 9 What Management Must Know About Forecasting ..........................................175

Aphorism 1: Forecasting Is a Huge Waste of Management Time 175

xii ▸  C O N T E N T S

Aphorism 2: Accuracy Is Determined More by the Nature of the Behavior Being Forecast than by the Specifi c Method Being Used to Forecast It 177

Aphorism 3: Organizational Policies and Politics Can Have a Signifi cant Impact on Forecasting Effectiveness 179

Aphorism 4: You May Not Control the Accuracy Achieved, But You Can Control the Process Used and the Resources You Invest 180

Aphorism 5: The Surest Way to Get a Better Forecast Is to Make the Demand Forecastable 182

Aphorism 6: Minimize the Organization’s Reliance on Forecasting 183

Aphorism 7: Before Investing in a New System or Process, Put It to the Test 184

Notes 185

Epilogue 187

Glossary 189

Appendix Forecasting FAQs 193Accuracy Expectations 193

Performance Benchmarks 196

Performance Measurement and Reporting 198

The Naïve Forecast 208

Forecast Value Added Analysis 211

Forecast Modeling 220

Politics and Practices of Forecasting 224

Demand Volatility 227

Forecasting Process 230

Judgment 237

Forecasting Organization 238

Low Volume/Intermittent Demand 239

New Product Forecasting 241

Forecasting Hierarchy 242

Software Selection 245

Index 247

xiii

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 soft-

ware 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

Foreword Tom Wallace

xiv ▸  F O R E W O R D

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 fi eld 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.

T OM W ALLACE

Sedona, Arizona

xv

We do not live in a boring world.

Companies do not operate the same, markets do not respond the

same way, and all customers defi nitely do not want the same things.

It ’ s this inexhaustible variability — and the careful blending of histori-

cal 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 fi nd 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 con-

tributions 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 vola-

tility 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 fi nal

forecast.

The accuracies of our forecasts across product lines, however,

varied greatly. Given the complex nature of our business, this was

Foreword Anne G. Robinson

xvi ▸  F O R E W O R D

expected. The true value of the forecasting process came from its

relative contribution against doing nothing. Based on Mike ’ s experi-

ences, 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.

A NNE G. R OBINSON , P H D

Sr. Manager, Information and Data Strategy

Customer Value Chain Management, Cisco

xvii

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 Good-

night 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 fore-

casting 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 Woodfi eld, 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.

Acknowledgments

xviii ▸  A C K N O W L E D G M E N T S

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 pre-

viously 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 C ô t é 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 mate-

rials, and to Rob Miller of Covidien for use of his “ Comet Chart. ”

Several of my SAS colleagues have been involved in the produc-

tion 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 Buffi e

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 .

A C K N O W L E D G M E N T S  ◂ xix

Special thanks go to a number of individuals who have had par-

ticular infl uence on my forecasting career, or on this book:

� John LaBella, VP — Application Delivery at Gap, Inc. John

gave me my fi rst 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 unoffi cial (and unpaid!) “ publicist ” for the

many writing, speaking, and professional association connec-

tions 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 fore-

casting 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 fi nished book. In addition, with her past

publishing experience, Meredith helped me survive the annoy-

ances 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 defi nitely succeed as a hostage

negotiator if this SAS gig ever falls through.

xx ▸  A C K N O W L E D G M E N T S

� Jessica Crews, Graphic Designer. Jessica brought to life the dis-

turbing 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 lead-

ership 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 fi eld of Sales &

Operations Planning, and has made major contributions world-

wide 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 fi elds would agree that Tom is just as much a celeb-

rity 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 valu-

able 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 refl ect 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.

Forecasting is a huge waste of management time.

The Business Forecasting Deal 1 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 benefi t achieved (the improvement in

accuracy).

We spend far too many organizational resources creating our

forecasts, while almost invariably failing to achieve the level of accu-

racy 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 effi ciency 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 fore-

caster 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 sim-

plest 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 deter-

mined 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.

Prologue

1

2 ▸  P R O L O G U E

The reality is that smooth, stable, repeating patterns can be fore-

cast 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 pro-

portional 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 accu-

racy 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 con-

taminate 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 ineffi ciency in the typical business fore-

casting 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 fi rm selling forecasts:

The goal of your organization is not accurate forecasts — it is to make a

profi t 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

benefi ts), 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, nonforecast-

ing, approaches that may more effectively solve the underlying busi-

ness problem.

P R O L O G U E  ◂ 3

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 unforecast-

able. 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 organiza-

tion more effi ciently and more profi tably. In some situations forecast-

ing 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 dis-

courage forecasting practitioners or to stifl e 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 fi ction.

4 ▸  P R O L O G U E

In writing The Business Forecasting Deal , I ’ ve pieced together self -

contained sections addressing specifi c 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 ineffi ciency in its fore-

casting process, and has been adopted at a number of major corpora-

tions. Some of these adopters have spoken publicly of their fi ndings,

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 specifi c 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

1. The Business Forecasting Deal is intended for both the forecasting practitioner and for management overseeing (or dealing with) the forecasting function. Sometimes ter-minology or formulas may appear that are not familiar to the reader, and are not thoroughly defi ned 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 fi rst time they appear in the text).

2. For a major infl uence 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.

Forecasting is a diffi cult, thankless, and sometimes futile endeavor.

When accuracy is not quite where everyone wants it to be, we

react by making signifi cant new investment in technology, process,

and people to solve the business problem. 1 While we live in an uncer-

tain 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 fi nite 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.

C H A P T E R 1

Fundamental Issues in Business Forecasting

5

6 ▸  F U N D A M E N T A L I S S U E S I N B U S I N E S S F O R E C A S T I N G

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 justifi cation is there for this belief that bread

nourishes? Just because bread has exhibited this nourishment char-

acteristic 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 justifi cation 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 con-

sistently 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.