OPERATIONS MANAGEMENT materials are used for multiple

products or if a product or service is

FORECASTING FUNDAMENTALS demanded by a number of independent
Types of Forecasts sources.
• Economic forecasts - Predict a variety
of economic indicators, like money 4. Forecast accuracy decreases as the
supply, inflation rates, interest rates, time period covered by the forecast—
etc. the time horizon—increases. Generally
• Technological forecasts - Predict rates speaking, short-range forecasts must
of technological progress and contend with fewer uncertainties than
innovation. longer-range forecasts, so they tend to
• Demand forecasts - Predict the future be more accurate.
demand for a company’s products or
services. ELEMENTS OF GOOD FORECAST
1. The forecast should be timely.
Forecasting Time Horizon 2. The forecast should be accurate,
Short Range Forecast. This forecast has a 3. The forecast should be reliable
time span of up to 1year but is generally less 4. The forecast should be expressed in
than 3 months. It is used for planning for meaningful units
purchasing, job scheduling, workforce levels, 5. The forecast should be in writing
job assignment and production levels. 6. The forecasting technique should be
simple to understand and use.
Medium Range Forecast. A medium range, or 7. The forecast should be cost-effective
intermediate, forecast generally spans from 3
months to 3 years. It is useful in sales STEPS IN THE FORECASTING PROCESS
planning, production planning and budgeting, - Determine the purpose of the forecast.
cash budgeting, and analysis of various - Establish a time horizon.
operating plans. - Obtain, clean, and analyze appropriate
data.
Long Range Forecast. Generally 3 years or - Select a forecasting technique
more in time span, long range forecasts are - Make the forecast.
used in planning for new products, capital - Monitor the forecast errors.
expenditures, facility location or expansion,
and research and development TYPES OF FORECASTING METHODS

FEATURES COMMON TO ALL FORECASTS Qualitative methods

1. Forecasting techniques generally These types of forecasting methods are based
assume that the same underlying on judgments, opinions, intuition, emotions, or
causal system that existed in the past personal experiences and are subjective in
will continue to exist in the future. nature. They do not rely on any rigorous
mathematical computations.
2. Forecasts are not perfect; actual results
usually differ from predicted values; the Quantitative methods
presence of randomness precludes a These types of forecasting methods are based
perfect forecast. Allowances should be on mathematical (quantitative) models, and
made for forecast errors. are objective in nature. They rely heavily on
mathematical computations.
3. Forecasts for groups of items tend to be
more accurate than forecasts for
individual items because forecasting
errors among items in a group usually
have a canceling effect. Opportunities
for grouping may arise if parts or raw
QUALITATIVE FORECASTING METHODS

Consumer Surveys
Because it is the consumers who ultimately
determine demand, it seems natural to solicit
input from them. In some instances, every
customer or potential customer can be
contacted. However, usually there are too
many customers or there is no way to identify
all potential customers. Therefore,
organizations seeking consumer input usually
resort to consumer surveys, which enable
QUALITATIVE FORECASTS them to sample consumer opinions. The
Executive Opinions obvious advantage of consumer surveys is that
- A small group of upper-level managers they can tap information that might not be
(e.g., in marketing, operations, and available elsewhere. On the other hand, a
finance) may meet and collectively considerable amount of knowledge and skill is
develop a forecast. required to construct a survey, administer it,
- This approach is often used as a part of and correctly interpret the results for valid
long-range planning and new product information. Surveys can be expensive and
development. time-consuming.
- It has the advantage of bringing together
the considerable knowledge and talents
of various managers. Delphi Method
- However, there is the risk that the view - Delphi method, an iterative process
of one person will prevail, and the intended to achieve a consensus
possibility that diffusing responsibility forecast. This method involves
for the forecast over the entire group circulating a series of questionnaires
may result in less pressure to produce a among individuals who possess the
good forecast. knowledge and ability to contribute
meaningfully.
Salesforce Opinions - Responses are kept anonymous, which
Members of the sales staff or the customer tends to encourage honest responses
service staff are often good sources of and reduces the risk that one person’s
information because of their direct contact opinion will prevail. Each new
with consumers. They are often aware of any questionnaire is developed using the
plans the customers may be considering for information extracted from the previous
the future. There are, however, several one, thus enlarging the scope of
drawbacks to using salesforce opinions. One information on which participants can
is that staff members may be unable to base their judgments.
distinguish between what customers
would like to do and what they actually
will do. Another is that these people are
sometimes overly influenced by recent
experiences. Thus, after several periods of
low sales, their estimates may tend to
become pessimistic. After several periods
of good sales, they may tend to be too
optimistic. In addition, if forecasts are
used to establish sales quotas, there will
be a conflict of interest because it is to the
salesperson’s advantage to provide low
sales estimates.
FORECASTS BASED ON TIME-SERIES DATA APPROACHES TO FORECASTING
- A time series is a time-ordered There are two general approaches to
sequence of observations taken at forecasting: Qualitative and Quantitative.
regular intervals (e.g., hourly, daily,
weekly, monthly, quarterly, annually). Qualitative methods consist mainly of
- The data may be measurements of subjective inputs, which often defy precise
demand, earnings, profits, shipments, numerical description.
accidents, output, precipitation, Quantitative methods involve either the
productivity, or the consumer price projection of historical data or the
index. Forecasting techniques based on development of associative models that
time-series data are made on the attempt to utilize causal (explanatory)
assumption that future values of the variables to make a forecast.
series can be estimated from past
values. Although no attempt is made to Qualitative techniques permit inclusion of
identify variables that influence the soft information (e.g., human factors, personal
series, these methods are widely used, opinions, hunches) in the forecasting process.
often with quite satisfactory results. Those factors are often omitted or downplayed
when quantitative techniques are used
1. Trend refers to a long-term upward or because they are difficult or impossible to
downward movement in the data. quantify.
Population shifts, changing incomes,
and cultural changes often account for Quantitative techniques consist mainly of
such movements. analyzing objective, or hard, data. They usually
avoid personal biases that sometimes
2. Seasonality refers to short-term, fairly contaminate qualitative methods. In practice,
regular variations generally related to either approach or a combination of both
factors such as the calendar or time of approaches might be used to develop a
day. Restaurants, supermarkets, and forecast.
theaters experience weekly and even
daily “seasonal” variations. Judgmental forecasts rely on analysis of
subjective inputs obtained from various
3. Cycles are wavelike variations of more sources, such as consumer surveys, the sales
than one year’s duration. These are staff, managers and executives, and panels of
often related to a variety of economic, experts.
political, and even agricultural Quite frequently, these sources provide
conditions. insights that are not otherwise available.

4. Irregular variations are due to unusual Time-series forecasts simply attempt to

circumstances such as severe weather project past experience into the future. These
conditions, strikes, or a major change in techniques use historical data with the
a product or service. They do not reflect assumption that the future will be like the past.
typical behavior, and their inclusion in Some models merely attempt to smooth out
the series can distort the overall random variations in historical data; others
picture. Whenever possible, these attempt to identify specific patterns in the data
should be identified and removed from and project or extrapolate those patterns into
the data. the future, without trying to identify causes of
the patterns.
5. Random variations are residual
variations that remain after all other Associative models use equations that
behaviors have been accounted for. consist of one or more explanatory variables
that can be used to predict demand. For
example, demand for paint might be related to
variables such as the price per gallon and the
amount spent on advertising, as well as to
specific characteristics of the paint (e.g.,
drying time, ease of cleanup).

NAÏVE METHODS
- Naive forecast uses a single previous
value of a time series as the basis of a
forecast.
- The naive approach can be used with a
stable series (variations around an
average), with seasonal variations, or
with trend. With a stable series, the last
data point becomes the forecast for the
next period. Thus, if demand for a
product last week was 20 cases, the
forecast for this week is 20 cases. With
seasonal variations, the forecast for
this “season” is equal to the value of
the series last “season”.

Techniques for Averaging

1. Moving average
2. Weighted moving average
3. Exponential smoothing

Moving Average
One weakness of the naive method is that the
forecast just traces the actual data, with a lag
of one period; it does not smooth at all. But by
expanding the amount of historical data a
forecast is based on, this difficulty can be
overcome. A moving average fore- cast uses a
number of the most recent actual data values
in generating a forecast. The moving average
forecast can be computed using the following
equation:
 the first three values as a forecast for period 4
Weighted Moving Average would provide a better starting forecast
A weighted average is similar to a moving because that would tend to be more
average, except that it typically assigns more representative.
weight to the most recent values in a time
series. For instance, the most recent value FORECAST ACCURACY
might be assigned a weight of .40, the next Accuracy and control of forecasts is a vital
most recent value a weight of .30, the next aspect of forecasting, so forecasters want to
after that a weight of .20, and the next after minimize forecast errors. However, the
that a weight of .10. Note that the weights complex nature of most real-world variables
must sum to 1.00, and that the heaviest makes it almost impossible to correctly
weights are assigned to the most recent value predict future values of those variables on a
regular basis. Moreover, because random
variation is always present, there will always
be some residual error, even if all other factors
have been accounted for. Consequently, it is
important to include an indication of the extent
to which the forecast might deviate from the
value of the variable that actually occurs. This
will provide the forecast user with a better
perspective on how far off a forecast might be.

Exponential Smoothing
Exponential smoothing is a sophisticated
weighted averaging method that is still
relatively easy to use and understand. Each
new forecast is based on the previous
forecast plus a percentage of the difference
between that forecast and the actual value of
the series at that point. That is:
Summarizing Forecast Accuracy
Forecast accuracy is a significant factor when
deciding among forecasting alternatives.
Accuracy is based on the historical error
performance of a forecast.

Three commonly used measures for

summarizing historical errors are the mean
absolute deviation (MAD), the mean
squared error (MSE), and the mean absolute
NOTE: percent error (MAPE).
Exponential smoothing should begin several - MAD is the average absolute error
periods back to enable forecasts to adjust to - MSE is the average of squared errors,
the data, instead of starting one period back. A and
number of different approaches can be used to - MAPE is the average absolute percent
obtain a starting forecast, such as the average error.
of the first several periods, a subjective The formulas used to compute MAD, MSE, and
estimate, or the first actual value as the MAPE are as follows:
forecast for period 2 (i.e., the naive approach).
For simplicity, the naive approach is used in
this book. In practice, using an average of, say,