DEMAND

FORECASTING
 DEMAND FORECASTING MEANS
PREDICTING OR ESTIMATING THE
FUTURE DEMAND FOR A PRODUCT WITH
A SCIENTIFIC APPROACH .
 IT IS UNDERTAKEN FOR THE PURPOSE
OF PLANNING AND MAKING LONGTERM
DECISIONS
Business Decision Making –Use of Demand Forecasting

 Planning and Scheduling

 Acquiring Inputs and make production
for future
 Making provision for finance i.e. capital
planning & capacity utilization
 Formulating pricing strategy
 Planning advertisement & sales
channels
Steps in Demand Forecasting

 Specifying the Objective

 Determining the time Perspective and
type of good
 Selecting a proper method of forecasting
 Collection of data
 Interpretation of results
Forecasting Horizons.
 Short Term (0 to 3 months): for inventory
management and scheduling.
 Medium Term (3 months to 2 years): for
production planning, purchasing, and
distribution.
 Long Term (2 years and more): for
capacity planning, facility location, and
strategic planning.
Factors involved in Demand
Forecasting
1. Undertaken at three levels:
a. Macro-level
b. Industry level
c. Firm level
2. Should the forecast be general or specific (product-
wise)?
3. Problems or methods of forecasting for “new” vis-à-vis
“well established” products.
4. Classification of products
5. Special factors peculiar to the product and the market –
risk and uncertainty.
METHODS OF DEMAND
FORECASTING
SURVEY METHODS
SURVEY
METHOS

CONSUMER OPINION
SURVEY METHODS
METHODS

COMPLETE SAMPLE EXPERTS TEST

END USE
ENUMERATION SURVEY OPINION MARKETING
METHOD
METHOD METHOD METHOD METHOD

DELPHI
METHOD
 STATISTICAL
METHODS

BAROMETRIC
REGRESSION METHOD
TRENDPROJECTION
METHODS
Consumer Survey
 Direct interview of potential consumer
( depending on purpose, time & cost in
research) by
 Complete enumeration method
 Sample survey method
 The end-use method
Consumer Survey (contd.)
 Complete enumeration method
 In this method, almost all the potential user’s are
contacted and are asked about their future plan of
purchasing the product in question.
 Dp = q1 +q2 +q3 +….+ qn = n summation (i=1)*qi
Consumer Survey (contd.)
 Sample survey
 Used when population of the target market is
very large.
 Only a sample of potential consumers or
users is selected for interview.
 Method can be direct interview or
questionnaire to the sample consumers.
 Dp = Hr / Hs ( H * Ad)
Consumer Survey (contd.)
 The end-use method
 Forecasting demand for inputs
 Method requires building up a schedule of
probable aggregate future demand for inputs
by consuming industries and other sectors.
 Technological, structural and other changes
are taken into consideration.
 The list of end users, norms of consumption,
Purpose, product wise or use wise demand
has to be assessed to make forecast.
OPINION POLL METHOD
(CONTD.)
 Nominal group technique
 Generation of ideas
 Collection of ideas
 Discussion
 Preliminary voting
 Final voting
OPINION POLL METHOD
 Expert opinion method
 Especiallysales representatives or
professional market experts or consultants
 Estimates may be restricted to their experts
and subjective judgment
OPINION POLL METHOD
(CONTD.)
 Delphi method: it consists of an effort
to arrive at a consensus in an uncertain
area by questioning a group of experts
repeatedly until the results appear to
converge along a single line of the
issues causing disagreement are
clearly defined.
OPINION POLL METHOD
(CONTD.) – Delphi Method
 Used to consolidate the divergent expert
opinions and to arrive at a compromise
estimate of future demand.
 Steps
 Selection of group of experts
 Ideas and forecasts are obtained

 Results are summarized and redistributed along

new questions
 Response again summarized
Delphi method
Advantages
1. Facilitates the maintenance of anonymity of the
respondent’s identity throughout the course.
2. Saves time and other resources in approaching a large
number of experts for their views.
Limitations/presumptions:
3. Panelists must be rich in their expertise, possess wide
knowledge and experience of the subject .
4. Presupposes that its conductors are objective in their
job, possess ample abilities to conceptualize the
problems for discussion, generate considerable thinking,
stimulate dialogue among panelists and make inferential
analysis.
Market Experiment method
 Consumer behaviour & market is studied
under actual but controlled market
conditions
 Selection of representative markets on
demographic similarities
 Market Experimentation by changing
controlled variables -prices, advertisement
expenditures etc.
 Consequent changes in demand are
recorded over a period of time
 Elasticity coefficients are computed &
implemented to assess future demand
 This method is cheaper than consumer
clinics & controlled laboratory experiments
 Fit for bigger concerns.
Statistical Methods
 Statistical methods are considered to be
superior due to the following reasons :

 The element of subjectivity is minimum

 Method of estimation is Scientific.

 Estimates are more reliable

Simple Moving Average
 Forecasts on the basis of demand values
during the recent past.
Month Order
Jan 120
Feb 180
Mar 200
Apr 100
May 110
Jun 50
Jul 80
Aug 120
Sept 150
Oct 90
Nov 120
Dec ?
Exponential smoothing
 Based upon the premise that the
importance of data diminishes as the past
becomes more distant
TIME SERIES PREDICTS
 This method uses historical and cross –
sectional data for estimating demand
 Finding a Trend value for a specific year

 FINDING SEASONAL FLUCTUATIONS IN

THE VARIABLE
 PREDICTING TURNING POINTS IN
FUTURE MOVEMENTS OF THE
VARIABLE
Analysis of time series and trend
projections

Four sets of factors: secular trend (T), seasonal

variation (S), cyclical fluctuations (C ),
irregular or random forces (I).
O (observations) = TSCI
Assumptions:
1. The analysis of movements would be in the
order of trend, seasonal variations and
cyclical changes.
2. Effects of each component are independent of
each other.
There are three techniques of trend
projection
 Graphical
 Fitting Trend Equation
 Box-Jenkins method
 The above method can be used by long
standing firms by using the data from
sales department and books of account .
 New firms can use older firms data
belonging to the same industry .
Linear Trend
It is represented: Y= a + b x (I)
 Y=Demand
 X= Time Period
 a & b are constants .
 For calculation of Y for any value of X
requires the∑Y=na+b∑X
values of a & b These are :

∑XY=a∑X+b∑X²
Problem & Solution
 The data relate to the sale of generator
sets of a company over the last five years

 Year : 2003 2004 2005 2006 2007

sets : 120 130 150 140 160

Estimate the demand for generator sets in

the year 2012 if the present trend
continues
 Year X Y x² Y² XY
2003 1 120 1 14400 120
2004 2 130 4 16900 260
 2005 3 150 9 22500 450
2006 4 140 16 19600 560
2007 5 160 25 25600 800
Total 15 700 55 99000 2190

Substituting table values in equation ii & iii we get

∑Y=na+b∑X
700 = 5a +15b
∑XY=a∑X+b∑X²
2190 = 15a +55b
By multiplying equation iv by 3 and subtracting it from equation v we get
10b =90
b =9
Solution
 Substitute this value in equation iv we have
 700 =5a +15 b
 700 = 5a +15 (9)
 5a =565
 a = 113
 Trend equation Y=113 + 9x
 For 2012 ,x will be 10
 Y2012 = 113+9 x 10 =203 sets
Simple Linear Regression
 Linear regression analysis establishes a
relationship between a dependent variable
and one or more independent variables.
 In simple linear regression analysis there
is only one independent variable.
 If the data is a time series, the
independent variable is the time period.
 The dependent variable is whatever we
wish to forecast.
Simple Linear Regression
 Regression Equation
This model is of the form:
Y = a + bX
Y = dependent variable
X = independent variable
a = y-axis intercept
b = slope of regression line
Simple Linear Regression
 Once the a and b values are computed, a
future value of X can be entered into the
regression equation and a corresponding
value of Y (the forecast) can be calculated.
Problem :
 The data of a firm relating to sales and
advertisement is given below .If the
manager decides to spend Rs 30 mill in
the year 2005 what will be the prediction
for sales
YEAR AD.EX SALES0 X2 XY
mill 000 units
1995 5 45 25 225
1996 8 50 64 400
1997 10 55 100 550
1998 12 58 144 696
1999 10 58 100 580
2000 15 72 225 1080
2001 18 70 324 1260
2002 20 85 400 1700
2003 21 78 441 1638
2004 25 85 625 2125
N=10 ∑X=144 ∑Y=656 ∑X2=2448 ∑XY=10254
 Solution
• a = (∑X²) ( ∑Y) - (∑X )( ∑X Y)

N∑X² - (∑X )²

b= N∑X Y - (∑X )( ∑ Y)
N∑X² - (∑X )²
a =(2448) (656)- (144)(10254)

10 (2448) - (144)2

= 1605888 - 1476576
= 129312
= 34.54
24480 - 20736 3744
 b= Value
• b= 10(10254)-(144)(656)

10 (2448) -(144)2
= 102540 -94464

24480 -20736

= 8076
= 2.15 THERE FOR : Y =a + b x
3744
Y=34.54 +2.15 x , x =30