Forecasting is the process of predicting changing conditions and future
events that may significantly affect the
business of an organization.
1. Forecasting is important to both planning and decision making.
2. Forecasting is used in a variety of
areas such as: production planning,
budgeting, strategic planning, sales
analysis, inventory control, marketing
planning and purchasing among others.
It’s important to look at forecasting effectiveness. Forecasting
techniques are most accurate when the
environment is not rapidly changing.
Some suggestions for improving forecasting effectiveness are as follows:
1) Use simple forecasting techniques.
2) Compare every forecast with “no change.”
3) Don’t rely on a single forecasting method.
4) Don’t assume that you can accurately identify turning points in a
5) Shorten the length of the forecasts.
6) Forecasting is a managerial skill and can
be practiced and improved.
Methods of Forecasting:
Quantitative forecasting relies
on numerical data and mathematical model
to predict future conditions. There are
two types of quantitative forecasting
most frequently used.
Time-series methods used
historical data to develop forecasts of
The underlying assumption is that patterns exist and that the
future will resemble the past.
Time-series methods do not in themselves predict the impact
of present or future actions that
managers might take to bring about
A trend reflects a long-range general movement is either an
upward or a downward direction.
A seasonal pattern indicates upward or downward changes that
coincide with particular points within a
A cyclical pattern involves changes at particular points in
time that span longer than a year.
Time-series are more valuable for predicting broad
environmental factors than in predicting
the impact of present or future actions.
Because time-series rely on past trends there can be a danger
in their use if environmental changes
Explanatory or causal models
attempt to identify the major
variables that are related to or have
caused particular past conditions and
then use current measures of those
variables (predictors) to predict future
Explanatory models allow managers to assess the probable
impact of changes in the predictors.
are equations that express the
fluctuations in the variable being
forecasted in terms of fluctuations
among one or more other variables.
are systems of simultaneous multiple
regression equations involving several
predictor variables used to identify and
measure relationships or
interrelationships that exist in the
are variables that tend to be
correlate with the phenomenon of major
interest but also tend to occur in
advance of the phenomenon.
Technological or Qualitative
Forecasting is aimed primarily at
predicting long-term trends in
technology and other important aspects
of the environment
The focus is upon longer-term issues that are less amenable to numerical
analysis as quantitative approaches.
The Delphi method and Scenario analysis can be used as techniques.
Judgmental Forecasting relies
mainly on individual judgments or
committee agreements regarding future
Judgmental forecasting methods are highly susceptible to
The jury of executive opinion is one of the two
judgmental forecasting model. It is a
means of forecasting in which
organization executives hold a meeting
and estimate, as a group, a forecast for
a particular item.
The Sales-force composite is a means of forecasting
that is used mainly to predict future
sales and typically involves obtaining
the views of various salespeople, sales
managers, and/or distributors regarding
the sales outlook.
The choice of which forecasting method to use depends upon the needs
within particular forecasting
1. Quantitative forecasting methods:
have a short-to-medium time horizon
require a short period of time if a method is developed
often have high development costs
are high in accuracy in identifying patterns
Are low in accuracy in predicting turning points for time
series, but medium for other methods.
Are difficult to understand
2. Technological forecasting methods:
have a medium-to-long time horizon
require a medium-to-long time
have medium development costs
are of medium accuracy in identifying patterns
are of medium accuracy in predicting turning points
Are easily understood.
3. Judgmental forecasting methods:
have a short-to-long time horizon
require a short time
have low development costs
are of medium-to-high accuracy in identifying patterns
are of low accuracy in predicting turning points
are easily understood
Download pdf file of Forecasting