UNIT_III

DEMAND ANALYSIS AND FORECASTING

Demand is crucial for the survival of any business enterprise. A firms own profit and/or sales,
depend mainly upon the demand for its product. A managements decisions on production,
advertising, cost allocation, pricing, inventory holdings, etc. all requires an analysis of demand.
Demand analysis attempts to identify and measure the factors that determine sales, on the basis
of which alternative methods of manipulating or managing demand can be worked out. Demand
forecasting attempts to estimate the expected future demand for a product, which helps to plan
production better. In this context, it is important to understand the types and determinants of
demand and their relative importance.
Demand is broadly classified as: (a) Demand for consumers goods and producers goods,
(b) Demand for perishable and durable goods,
(c) Derived and autonomous demands,
(d) Firm and industry demands, and
(e) Demand by total market and by market segments.
a) Consumers goods and producers goods - Consumers goods are directly used for final
consumption. Meanwhile, producers goods are used for further production of other goods, which
may either be in the form of consumers or producers goods. The former includes clothes,
houses, food, etc., while the latter includes machines, tools, raw materials, etc. Consumers goods
are also known as direct demand. Whereas producers goods is known as derived demand.
b) Perishable and durable goods demand - Consumers and producers goods are further
classified as perishable and non-durable goods. Those goods which can be consumed only once
are known as perishable goods, whereas the durable goods can be used more than once during a
period of time. For example, vegetables, fruits and milk are perishable consumer goods, while
oil, raw materials and coal are non-durable producer goods. On the other hand, car, refrigerator
and furniture are durable consumers goods, while industrial buildings, machine and tools are
durable producers goods.
c) Derived and autonomous demand When the demand for a good is associated with another
parent good, it is called derived demand. For example, the demand for steel is not for its own
sake, but for satisfying the demand for construction. In this sense, the demand for all producers
goods is derived. On the other hand, autonomous demand is wholly independent of all other
demands. It is difficult to name a product which is fully autonomous
d) Firm and industry demands Firm demand represents the demand for products of a single
company, while industry demand refers to the demand of an industry.
e) Demands by total market and by market segments The total market demand for a product
refers to the total demand, while the demand arising from different segments of the market is
market segment demand. Segments include different regions, product use, distribution channels,
customer sizes, and sub-products. Each of them differ significantly with respect to delivered
prices, net profit margins, competition, seasonal patterns and cyclical sensitivity. Wide

differences in them call for a demand analysis restricted to an individual market segment, which
in turn would help a firm to manipulate the total demand. Hence, a A company/ industry would
be interested in the both these demands. Risk and uncertainty are involved in every decisionmaking process. The producer, manager or any decision-making authority should be aware of the
existing level of demand for the products being produced, and estimate the gap between demand
and supply. In a growth-oriented decision-making process, the manager decision-maker is
expected to know the changes that are expected to take place in the future demand. Such
knowledge would help to determine the targets to be achieved to match the future demand with
the available supply. Thus, the manager decision-maker, whether a firm or a state planning
agency, must not only estimate the present level of demand, but should also forecast the future
demand (Barla 2000).
The extent of objectivity and precision with which demand for a product is estimated and
projected for the future would determine the ability of a decision-making agent in dealing with
further uncertainties. For example, if there is a possibility of rise in the prices of petroleum
products, the automobile producers may plan to switch over to the production of smaller cars.
Such switch-over decisions need to be made on the basis of accuracy of demand forecasts. Thus,
major decisions in business enterprises depend upon forecasts of one kind or the other.
Stages in forecasting demand
Based on the scope of demand forecasting for a commodity, the following sequence is generally
adopted in projecting demand: (1) Specification of objective(s): Specification of the purpose of demand forecasts is the
foremost task in forecasting demand.
(2) Selection of appropriate technique: Next, selection of appropriate technique for the purpose
is important. If it is proposed to use regression method, the model has to be specified properly by
identifying the necessary variables and the nature of relationship between X and Yj.
(3) Collection of appropriate data: Collection of quality and adequate data for the demand
forecasting would determine the quality and reliability of results. Hence, the data collected
should also be representative.
(4) Estimation and interpretation of results: The results obtained through the analysis of
collected data, either manually or with the help of computers, should be interpreted carefully in
correspondence with the objectives examined.
(5) Evaluation of the forecasts: A model used for demand forecasting with objectivity, would
yield good results. The results, however, need to be verified by persons possessing professional
acumen and expertise.
Data and techniques of demand forecasting
A good set of data is required for the estimation of present level of demand and forecasting the
future demand. A private sector forecasts demand on the basis of past experience and the data
collected from various sources. Similarly, a public sector uses data collected by different
government and research agencies for the purpose. The following are some of the techniques
adopted for estimating the existing and future demands.

Demand Forecasting is the method of predicting the future demand for the firms
product. It is guess or anticipation or prediction of what is likely to happen in the
future. Forecast can be done for several things. It is based on the experience.

Techniques or methods of Demand Forecasting: Method of Demand Forecasting is based on whether

the good is Established Good or new good.

Methods of Demand Forecasting for established goods:

Information of the established good is available so the forecast can be based on this information. Two
basic methods of demand forecasting for the established goods are:

Interview and Survey Approach (for short period forecast): Interview and Survey Approach
collects information in the different way. Depending upon how the information is collected, we
have different sub methods as follows:

Opinion-Polling Method: This method tries to collect information from the customer
directly or indirectly through market research department of the firm or through the
whole sellers or the retailers. Consumers are contacted through mails or phones or
Internet and information regarding their expected expenditure is collected. This
method is useful when consumers are small in number.
Limitations:

It is difficult and costly to contact all the customers

It is suitable only for short period
Consumers are not sure of their purchase plans

Collective Opinion Method: Large firms have organized sales department. The
salesman has the technical training as how to collect the information from the buyers.
This information is further used for forecasting the demand.
Limitations:

It is difficult and costly to contact all the customers

It is suitable only for short period
This is based on judgment & has no scientific basis.

Sample Survey Method: The total number of consumers for the firms product is very
large called as population. It is practically not possible to contact all the consumers.

Only few of them are contacted and this forms the sample. The sample forecasts are
then generalized for the whole population through advanced statistical methods
available.
Limitations:

Information collected may not be accurate.

Sample is not a random sample.
Consumers do not have the correct idea of their purchases in future.

Panel of experts: Panel of experts consists of persons either from within the firm or
from outside the firm. These experts come together and forecast the demand for their
product that is purely based on the judgment of these experts so they are less
accurate. But if based on the scientific method the forecast would be accurate.
Composite management opinion: The opinions of the experienced person within the
firm are collected and manger analyses this information. This method is quick, easy
and saves time, but is not based on the scientific analysis and thus may not give very
accurate results.
Projection Approach(for long period forecast): In this method past experience is projected
into the future. This can be done with the help of statistical methods.

Correlation and Regression Analysis: Past data regarding the factors affecting the
demand can be collected. It is possible to express this on the graph. This is a scatter
diagram.
Example: If we collect the past data about the sales and advertising expenditure of the
firm,
it is possible to express in the form of scatter diagram as shown below:

* *
*
*
* *
*
*
* *
*
* *
A

Sales

X
O

Advertisement Expenditure

In the above diagram we get the functional relationship as line AA. Here
Advertisement Expenditure is the independent variable and Sales is the dependent
variable. The relationship between these variables is correlation and the technique of
establishing this relationship is regression. In simple correlation we establish
relationship between 2 variables and more than 2 variables in multiple correlation.
Limitations:

Assumption made is that correlation between two variables will

continue in future also, this might not happen.

Time Series Analysis: Demand forecasts for a period of 2-3 years are based on time
series analysis. It is similar to the correlation analysis. It is based on the assumption
that the relationship between the dependent and the independent variable continues
to hold in the future.

Methods of Demand Forecasting for new products:

Indirect methods of forecasting are used to estimate demand for new products. Following are the
methods suggested:

Evolutionary Method: Some new goods evolve from already established goods. Demand
forecast for such new good is based on already established good from which they are evolved.
For example Demand for the color TV can be calculated from Demand for the black and white
TV, from which it is actually evolved.
Limitations:

The product should have been evolved from the existing product.
It ignores the problem of how the new product differs from the old
product.

Substitution Method: Some new goods are substituted of already established goods. For
example VCR substituted with VCD player.
Limitations:

New product may have many uses and each use has different
substitutability
When the substitute is added is added into market existing firm may react
by changing the prices.

Opinion Polling Method: Expected buyers and the consumers are directly contacted and
opinion about the product is directly taken from them. If the population is large then sample is
selected and results are generalized for the population.
Limitations:
It is difficult and costly to contact all the customers

It is suitable only for short period

Consumers are not sure of their purchase plans

Sample Survey Method: New product are first introduced in the sample market and the results
seen in the sample market are generalized for the total market.
Limitations:

Information collected may not be accurate

Tastes and the preferences may differ from market to market

Indirect Opinion Polling Method: Opinion of the consumers is indirectly collected through the
dealers who are aware of the needs of the customers.
Limitatios:

It is based on the judgment

Limited Scope

How is demand forecast determined?

There are two approaches to determine demand forecast (1) the qualitative approach, (2) the
quantitative approach. The comparison of these two approaches is shown below:
Quantitative Approach

Description

Qualitative Approach

Applicability

Considerations

Used when situation is stable &

Used when situation is vague & little historical data exist
data exist (e.g., new products and
technologies)
(e.g. existing products, current
technology)
Involves intuition and experience

Involves mathematical techniques

Jury of executive opinion

Time series models

Sales force composite

Causal models

Techniques
Delphi method
Consumer market survey
Qualitative Forecasting Methods
Your company may wish to try any of the qualitative forecasting methods below if you do not
have historical data on your products' sales.

Qualitative Method
Jury
of
opinion

Description

executive The opinions of a small group of high-level managers are

pooled and together they estimate demand. The group uses
their managerial experience, and in some cases, combines the
results of statistical models.

Sales force composite

Each salesperson (for example for a territorial coverage) is

asked to project their sales. Since the salesperson is the one
closest to the marketplace, he has the capacity to know what
the customer wants. These projections are then combined at
the municipal, provincial and regional levels.

Delphi method

A panel of experts is identified where an expert could be a

decision maker, an ordinary employee, or an industry expert.
Each of them will be asked individually for their estimate of
the demand. An iterative process is conducted until the
experts have reached a consensus.

Consumer

market The customers are asked about their purchasing plans and

survey

their projected buying behavior. A large number of

respondents is needed here to be able to generalize certain
results.

QUANTITATIVE FORECASTING METHODS

There are two forecasting models here (1) the time series model and (2) the causal model. A
time series is a s et of evenly spaced numerical data and is o btained by observing responses at
regular time periods. In the time series model , the forecast is based only on past values and
assumes that factors that influence the past, the present and the future sales of your products will
continue.
On the other hand, t he causal model uses a mathematical technique known as the regression
analysis that relates a dependent variable (for example, demand) to an independent variable (for
example, price, advertisement, etc.) in the form of a linear equation. The time series forecasting
methods are described below:

Description

Time
Series
Forecasting
Method
Assumes that demand in the next period is the same as demand in most
Nave Approach recent period; demand pattern may not always be that stable
For example:
If July sales were 50, then Augusts sales will also be 50

Description

Time
Series
Forecasting
Method

Moving
(MA)

MA is a series of arithmetic means and is used if little or no trend is present

Averages in the data; provides an overall impression of data over time
A simple moving average uses average demand for a fixed sequence of
periods and is good for stable demand with no pronounced behavioral
patterns.
Equation:
F 4 = [D 1 + D2 + D3] / 4

F forecast, D Demand, No. Period

(see illustrative example simple moving average)
A weighted moving average adjusts the moving average method to reflect
fluctuations more closely by assigning weights to the most recent data,
meaning, that the older data is usually less important. The weights are
based on intuition and lie between 0 and 1 for a total of 1.0
Equation:
WMA 4 = (W) (D3) + (W) (D2) + (W) (D1)
WMA Weighted moving average, W Weight, D Demand, No.
Period
(see illustrative example weighted moving average)
Exponential
Smoothing

The exponential smoothing is an averaging method that reacts more

strongly to recent changes in demand by assigning a smoothing constant to
the most recent data more strongly; useful if recent changes in data are the
results of actual change (e.g., seasonal pattern) instead of just random
fluctuations
F t + 1 = a D t + (1 - a ) F t

Where
F t + 1 = the forecast for the next period
D t = actual demand in the present period
F t = the previously determined forecast for the present period

 = a weighting factor referred to as the smoothing constant

(see illustrative example exponential smoothing)

The time series decomposition adjusts the seasonality by multiplying the

Time
Series normal forecast by a seasonal factor
Decomposition
(see illustrative example time series decomposition)

SIGNIFICANCE OF DEMAND FORECASTING

Estimating and forecasting demand are crucial to the following types of decision-makers for
knowing the present level of demand and the expected increase in demand over time.
(i) Producers: A producer allocates various factors of production for maximization of profit, for
which knowledge of both the present and future demand are important. Future demand estimates
helps the producer to plan the extent of expansion in scale of operations, so as to deal with the
increased demand and earn higher profits.
(ii) Policy makers and planners: It helps government to formulate economic policies through
the planning boards or planning commissions to allocate resources for economic development
through production in the public, private and export sectors to achieve the targets set for a given
time period. It also ensures adequate supply of inputs for achieving the objectives of industrial
policy, import-export policies, credit policy, public distribution system, and other related
policies, which involves forecasting of future demand.
(iii) Other groups of the society: Demand forecasts are also useful to researchers, social
workers and others with futuristic approach, to understand the levels of future demand or supply,
the gaps, and their expected impact on prices or the economy.