2. “Demand estimation (forecasting) may be defined as a process
of finding values for demand in future time periods.”-
According to Evan J. Douglas.
“The act of forecasting is of great benefit to all who take
part in the process and is the best means of ensuring
adaptability to changing circumstances. The collaboration of
all concerned lead to a unified front, an understanding of the
reasons for decisions and a broadened outlook”.- According
to Henry Fayol.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 2
3. Preparing the budget proposal
To Fulfilling objectives organization
Stabilizing production and employability
Future Expanding organizations
Taking correct Management Decisions
Evaluating Performance appraisal
Increasing government revue
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 3
4. Helpful in deciding to enter a new market or not.
Determining the sales territories.
Helpful in determining productivity.
Determining the product pricing strategy.
In decide the number of sales persons required to
achieve the sales objective.
Assessing the effect of a proposed marketing
programmed.
Product mix decisions.
Deciding the channels of Distribution
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 4
5. The following factors are to be considered for while going demand
forecasting:
Product Price
Purchasing power of customers
Demography
Replacement demand
Credit conditions
Conditions within the industry
Socio economic conditions.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 5
6. Methods or Techniques of Forecasting
The forecasting techniques generally classified two
types:
1. Qualitative methods
2. Quantitative methods.
Forecasting methods use mathematical model and
existing historical data and experiences of the veterans
or combination of these two.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 6
7. I. Qualitative or Judgmental Method
1. Sales force opinion
2. Executive opinion
3. Delphi Technique
4. Market Research
5. Customer survey
II. Quantitative or causal Method
1. Linear Regression
2. Time Series
3. Simple Average
4. Weighted Average
5. Exponential smoothing
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 7
8. I. Judgment or Qualitative Methods
Qualitative methods use the opinions of the veterans which are
translated in to action here no historical data are used only expert
opinions, survey reports and feedback from sales force are used to
estimate and forecast.
A) Sales Force Estimate
Sales force in their respective territories provides feedback on the sales
which are used estimate the expected sales. Many sales force feedbacks
are consider making companies overall sales estimate this would be
accurate as sale man would be in direct contact with customers.
Merits:
Can easily estimate product and services the customer would avail as
sales person would be direct contact with customers.
Forecasting for the new product will be very easier especially in
industrial market.
Demerits:
It is a completely subjective method.
The sales person may give the lower estimates if the estimates alone
are used to set their sales quotas.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 8
9. B) Executive opinion
Executive opinion would be used mostly on technical
fore casting or modify the existing forecast due to
unforced seen events. This may include the opinion from
a single person or more than one person.
Merits:
Forecast may be made quickly and economically.
Much more factual than made from consumer opinion
and sales force method.
Demerits:
It is very subjective and hence forecast lacks scientific
reality.
The executives may rate recent experiences more heavily
than more distant once which may result in too much
optimism or pessimism regarding future sales.
It takes more time and hence it will incur high expense.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 9
10. C) Market research
Market research uses methods for identifying customer interest. The
methods used can be market surveys which as series of questions are
answer by customers. Other methods could be.
Survey questionnaire covering economic and demographic information
from each person interviewed.
Gathering the customer interest by communicating through telephone
polling, mailings, or personal interviews.
Merits
The new entry products especially introduce for the first time.
Sales f
Demerits
Sales forecast data are projected on the basis of results of a part of the
segment or the market.
It takes long time to test the market.
orecast is based on actual results hence forecast is more reliable.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 10
11. D) Delphi method
The Delphi method is process were group of experts provide
their estimate based on the experience and logical thinking.
Generally this method is used to forecast for a long term or for
new product.
The group of experts will report individually to the
coordinator and give their opinions. Hence experts will not
know about the opinion provide by others. The coordinator
will logically take a decision to select the best opinions.
Merits
Changes in society.
Government regulations.
The competitive environment.
Demerits
There would be lot of opinions and coordinator may take a
long time to assess and select the best one.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 11
12. II. Quantitative Methods (or) casual methods:-
In this method huge historical data are used in
scientific way to analysis a future demand and
competitiveness of the products. Statistical methods
used are generally group under time series analysis.
Few of them are given below.
A)Regression Methods
In these methods various formulas are formulated and
formulas provided relationship between the variables
consider. This has many types. Few of them are given
below.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 12
13. (i)Linear regression
Linear regression is a mathematical technique that uses formulas similar to a straight
line. Here two variables one is independent and other one is dependents are related. This is
given as below.
y=a + b x
Where:
y = dependent variable
x = independent variable
a = y-intercept of the line
b = slope of the line.
(ii) Multiple Regression
When dependent variable has more than one independent variable than multiple regression
methods is used. Now the relationship between dependent variables and independent variable
is given as follows.
Y= A0+ A1 X1 +A2 X2 +………+ An Xn
Where
Y = dependent variable
A0 = the intercept
A1,… An = Parameter representing the contribution of independent variable
X1 ,…Xn= Independent variable
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 13
14. (b)Time series methods:-
Here for a given periods of time varies methods to
calculate future demand are formulated. Few of them
are as follows.
(i) Simple Moving Averages:
The demands for the given periods of time are
considered and their summation is divided by the no
of time periods considers would provide demand for
the next
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 14
15. (ii) Weighted Moving Averages:
Here for a given period of time T is divided into n number of periods the
demand for this periods are calculated as follows.
The weighted averages for say M number of periods are considered.
Demand for the first M number of periods are totaled and divided by M to
get the demand for M th periods.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 15
Periods(N) Sales Demands Periods(N) Sales
1. 31 - 1. 31
2. 35 - 2. 35
3. 33 (31+35+33)/3=33 3. 33
4. 40 (35+33+40)/3=36 4. 40
5. 32 (33+40+32)/3=35 5. 32
16. c) Exponential smoothing.
Here the exponential smoothing relationship exists
between the independent and dependent variables. This
given by.
Y=ex
This is further simplified
Y=C*X1+(1-C)*X2
Where
Y= future demand or next demand
X1= demand of the latest period
X2= demand of the last but one period.(previous periods of
last one).
C= smoothening constant and varies between zero and one.
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 16
18. Solution:
Let us four period average as the initial forecast year 5
while smoothing constant of C=0.1
S5=(S1+S2+S3+S4)
=(3+4+6+5)=18/4=4.5
S6= c*S5+(1-c)*Smt
S6 = 0.1*5+(1-0.1) 4.5
=4.55
Similarly: the sales for year 7 can be work out.
S6 = c*S6+(1-c)*Smt
= 0.1*8.6+(1-0.1)*4.55
=4.955
Mr.K.Vinothkumar, Ass Prof,
Mechanical, SRMIST, Ramapuram 18