International Journal of Forecasting 29 (2013) 525–526

Contents lists available at SciVerse ScienceDirect

International Journal of Forecasting

Book review

Fundamentals of Demand Planning and Forecasting, Part III (Chapters 11–13) is devoted to data: what you
Chamin L. Jain, Jack Malehorn. Graceway Publishing need to know, what to look for, how to treat it, and how
Company, Inc. (2012). 416 pp., softcover, $86.95, much to use.
ISBN: 978-0-9839413-0-9 Part IV (Chapter 14) summarizes the fundamentals
of modeling, with fifteen general rules (including: ac-
Jain and Malehorn’s recent book of forecasting and de- tual = pattern + error; a model can be improved further,
mand planning fundamentals is a comprehensive reference but at a cost; sophisticated models are not necessarily bet-
for beginners and experienced forecasting practitioners ter than simple ones).
alike. In a clear manner, the authors cover virtually every- Part V (Chapters 15–21) explains the theory behind
thing that a forecaster should know in order to be effective. the different types of time series models used by 62% of
In answer to the question, ‘‘Why this book?’’, Alan Milliken respondents to the Institute of Business Forecasting and
states in the foreword that the ‘purpose of this book is to Planning (IBF)’s 2009 survey (p. 133). Models are catego-
present the ‘‘Fundamentals of Demand Planning and Fore- rized as averages, moving averages, exponential smooth-
casting’’ to all in the firm who want to learn how to develop ing, trend line, classical decomposition, sales ratios and
family member forecasting. (The last is not generally con-
accurate forecasts, communicate them to decision makers,
sidered a forecasting method, but rather as a way of dis-
and use the output to gain a competitive advantage.’ With
aggregating a higher level forecast into its components
this purpose in mind, the book (ten years in the making, ac-
(family members), for example from product line to SKU,
cording to the authors) strives to cover all aspects of fore-
or SKU to SKU/location.) Each chapter explains the method
casting, and succeeds admirably!
clearly, using step-by-step, easy-to-follow examples.
The book is organized into thirty-one chapters, each fol-
Part VI (Chapters 22–25) focuses on cause-and-effect
lowed by Questions for review, and some also with Sug-
models. As you would expect, regression modeling is in-
gested readings, meaning that it could also be used as a
cluded here, but, despite the fact that they are normally
textbook. It also has an ample glossary of forecasting
classified with other univariate models, Box-Jenkins mod-
and supply chain terms, including formulas for calcu-
els are also covered here, because ‘‘Since they work very
lation where applicable. For example, for coefficient of
much like regression and the knowledge of regression is
variation, the definition ‘‘Measure used to determine the needed to understand them, we decided to discuss them
forecastability of a data series: The lower the COV, the eas- after the chapters on regression’’. (p. 240). The authors’
ier it should be to forecast. It is the percentage of variation explanation of Box-Jenkins models (including a section
around the arithmetic mean of a series, and is calculated as on ‘‘Things you should know about ARMIA modeling’’) is
follows:’’ is followed by the formula: coefficient of varia- one of the clearest and most easy-to-understand that this
tion = (standard deviation/mean) × 100. The appendices writer has seen anywhere! Chapter 25, on neural networks,
present simple instructions for computing the coefficient which was the only chapter written by an independent
of correlation and the standard deviation, and running a re- consultant, has less value for the typical forecasting prac-
gression model in Microsoft Excel 2007, plus the Student’s titioner, since these methods are not currently common
t and F distribution tables. An index of topics covered in among business forecasters. (Only 4% of the respondents
the book is also included. to the 2009 IBF survey who used cause and effect models
Part I (Chapters 1–5) details the basics of forecasting— (16% of the total) indicated that they used these methods;
what and why, data, models, placement of forecasting see pp. 133, 204.)
functions, and demand forecasting vs. supply planning. Part VII (Chapter 26) describes the most common per-
The forecasting process is covered in Part II (Chapters formance metrics (mean percent error (MPE), mean ab-
6–10), with a focus on how to build a successful process: solute percent error (MAPE), weighted mean absolute
requirements, ingredients, need for collaboration, support percent error (WMAPE) and bias), their meanings, and how
from top management, reaching a consensus, sales and they are calculated. It also addresses the use of the ‘‘range
operations planning (S&OP), and collaborative planning, of error’’, ways of improving forecasts, and the progress
forecasting and replenishment (CPFR). made in improving accuracy, based on the IBF surveys from
0169-2070/$ – see front matter
http://dx.doi.org/10.1016/j.ijforecast.2013.03.001
526 Book review / International Journal of Forecasting 29 (2013) 525–526

2000 to 2009 for all industries and consumer product in- in computer technology, software and statistical methods
dustries, comparing the two five-year periods, 2000–2004 as well as with the use of consumption data, demand plan-
and 2005–2009. ning and forecasting will improve’’. However, they expect
Part VIII (Chapter 27) addresses the important topic of that opposing forces as a result of globalization, more new
how to present, report and sell forecasts, with some key products, distribution channels and shorter product life cy-
findings from the personal experience of two gentlemen cles will counteract this.
in the pharmaceutical industry. One weakness of the survey results cited throughout
Part IX (Chapter 28) presents worst practices, the au- the book is that, though the book was published in 2012,
thors’ perspective on recent trends to the list of such the latest survey information was collected in 2009. One
practices started by Michael Gilliland in The business fore- wonders how the results might have differed had they
casting deal.1 See also Smith and Clarke (2012). They cat- been more timely. Further, although the IBF surveys of
egorize ‘‘worst practices’’ in misunderstandings of the attendees at their conferences are among the few regular
basics (e.g., forecasts are different to business plans), then surveys of forecasting practioners, one should keep in mind
group the worst practices within S&OP (e.g., multiple fore- that the results may not represent the general practitioner
casts within the company) and within CPFR (e.g., no agree- population, since the attendees represent a select group of
ment on the metrics to be used to evaluate performance), forecasters whose organizations have enough interest in
in addition to types of forecasts, model selection misper- improving their forecasting and a sufficient training/travel
ceptions (e.g., the best fitting models give the best fore- budget to send their employees to these conferences.
casts), metrics, and software packages, using a company Despite the above minor reservations, I heartily recom-
example and case studies. mend this book to all involved in the forecasting challenge
Good suggestions for the selection of a forecasting as one of the most complete and clearly written forecast-
software package and the development of a forecasting ing references available. The authors are to be commended
system are contained in Part X (Chapters 29–30). For for their success in realizing the purpose for this book, as
software, these include: spell out the requirements driven stated in the foreword.
by company needs (a list of considerations is included on
pp. 329–332); and package does not replace process. For References
system, suggestions include: before you start looking, you
need support from top management, approval of stake- Gilliland, M. (2010). The business forecasting deal. Hoboken, New Jersey:
John Wiley & Sons, Inc.
holders, and determination of what you’d like to have, Smith, J., & Clarke, S. (2012). Our best worst forecasting mistakes.
based on requirements and what is available. This sec- Foresight, 25, 16–20.
tion concludes with lessons learned in implementing and
maintaining a system at Merkel Inc. and other companies
(pp. 341–344).
Part XI (Chapter 31) contains the authors’ view of the Carolyn I. Allmon
future of demand planning and forecasting: the increasing Carolyn Allmon Business Forecasting Services,
role of collaboration, advances in technology, more statis- Eden Prairie, MN 55347,
tical analysis and less judgment, and an increased use of United States
online data. Their final assessment is that ‘‘with advances E-mail address: [email protected].

1 Gilliland (2010, pp. 29–79).