Operations management M103

Unit 2 Demand Management and Forecasting

Demand management

Demand management is the supply chain management process that balances the customers’ requirements
with the capabilities of the supply chain. With the right process in place, management can match supply
with demand proactively and execute the plan with minimal disruptions. The process is not limited to
forecasting. It includes synchronising supply and demand, increasing flexibility, and reducing variability.

This includes forecasting demand and synchronising it with production, procurement, and distribution
capabilities.

• A good demand management process can enable a company to be more proactive to anticipated
demand, and more reactive to unanticipated demand.

Understanding Demand Management

Demand management encompasses a range of activities aimed at aligning an organization’s supply chain
processes with customer demand. Its primary goal is to achieve a delicate balance between supply and
demand to optimize the use of resources and enhance customer satisfaction. Key aspects of demand
management include:

1. Demand Forecasting

Accurate demand forecasting is the cornerstone of demand management. Organizations use historical
data, market trends, and statistical models to predict future demand for their products or services.

2. Demand Planning

Once demand is forecasted, demand planning involves creating strategies and action plans to meet that
demand. It includes decisions about inventory levels, production schedules, and procurement.

3. Demand Shaping

Demand shaping involves influencing customer demand through marketing and pricing strategies. For
example, offering promotions or discounts during off-peak periods can help smooth demand.

4. Demand Sensing

Demand sensing refers to real-time monitoring of customer demand signals, such as point-of-sale data, to
respond quickly to changes in demand.

The Significance of Demand Management

Demand management plays a crucial role in supply chain operations for several reasons:
1. Efficient Resource Allocation

By accurately forecasting and planning for demand, organizations can optimize the allocation of resources,
reducing the risk of overstocking or stockouts.

2. Cost Reduction

Efficient demand management can lead to cost savings by minimizing excess inventory carrying costs and
production inefficiencies.

3. Customer Satisfaction

Meeting customer demand promptly and reliably enhances customer satisfaction and loyalty.

4. Enhanced Responsiveness

Effective demand sensing and shaping enable organizations to adapt quickly to changing market conditions
and customer preferences.

5. Risk Mitigation

Demand management helps in identifying and mitigating risks associated with supply chain disruptions,
ensuring business continuity.

Key Components of Demand Management

1. Data Analytics

Data analytics tools and techniques are essential for processing and analyzing large datasets to generate
accurate demand forecasts.

2. Collaboration

Collaboration among different departments, including sales, marketing, and supply chain, is crucial for
effective demand management.

3. Technology

Advanced software and technology solutions, such as demand planning software and Enterprise Resource
Planning (ERP) systems, facilitate demand management processes.

4. Continuous Improvement

Organizations should continuously review and refine their demand management processes to adapt to
changing market dynamics.

Challenges in Demand Management

Demand management is not without challenges, including:

1. Forecasting Accuracy

Achieving precise demand forecasts can be difficult, especially in dynamic markets with fluctuating
customer preferences.

Data Quality

The accuracy and quality of data used for forecasting are critical. Inaccurate data can lead to erroneous
forecasts.

3. Demand Variability

Managing demand for products with high demand variability can be challenging, as sudden spikes or drops
in demand are harder to predict and plan for.

4. Market Dynamics

External factors such as economic conditions, competition, and geopolitical e vents can impact demand and
require adaptive demand management strategies.

Demand Forecasting

Demand forecasting seeks to investigate and measure the forces that determine sales for existing and new
products. Generally companies plan their business – production or sales in anticipation of future demand.
Hence forecasting future demand becomes important. In fact it is the very soul of good business because
every business decision is based on some assumptions about the future whether right or wrong, implicit or
explicit. The art of successful business lies in avoiding or minimizing the risks involved as far as possible and
face the uncertainties in a most befitting manner .Thus Demand Forecasting refers to an estimation of most
likely future demand for a product under given conditions. Important features of demand forecasting It is
basically a guess work – but it is an educated and well thought out guesswork. It is in terms of specific
quantities It is undertaken in an uncertain atmosphere. A forecast is made for a specific period of time
which would be sufficient to take a decision and put it into action. It is based on historical information
and the past data. It tells us only the approximate demand for a product in the future. It is based on
certain assumptions. It cannot be 100% precise as it deals with future ex

Demand forecasting is needed to know whether the demand is subject to cyclical fluctuations or not, so
that the production and inventory policies, etc, can be suitably formulated Demand forecasting is generally
associated with forecasting sales and manipulating demand. A firm can make use of the sales forecasts
made by the industry as a powerful tool for formulating sales policy and sales strategy. They can become
action guides to select the course of action which will maximize the firm‟s earnings. When external
economic factors like the size of market, competitors attitudes, movement in prices, consumer tastes,
possibilities of new threats from substitute products etc, influence sales forecasting, internal factors like
money spent on advertising, pricing policy, product improvements, sales efforts etc., help in manipulating
demand. To use demand forecasting in an active rather than a passive way, management must recognize
the degree to which sales are a result not only of external economic environment but also of the action of
the company itself.

Managerial uses of demand forecasting:

In the short run: Demand forecasts for short periods are made on the assumption that the company has a
given production capacity and the period is too short to change the existing production capacity. Generally
it would be one year period.

 In Production planning: It helps in determining the level of output at various periods and avoiding
under or over production.
 Helps to formulate right purchase policy: It helps in better material management, of buying inputs
and control its inventory level which cuts down cost of operation.
 Helps to frame realistic pricing policy: A rational pricing policy can be formulated to suit short run
and seasonal variations in demand.
 Sales forecasting: It helps the company to set realistic sales targets for each individual salesman
and for the company as a whole.
 Helps in estimating short run financial requirements: It helps the company to plan the finances
required for achieving the production and sales targets. The company will be able to raise the
required finance well in advance at reasonable rates of interest.
 Reduce the dependence on chances: The firm would be able to plan its production properly and
face the challenges of competition efficiently.
 Helps to evolve a suitable labour policy: A proper sales and production policies help to determine
the exact number of labourers to be employed in the short run.

In the long run:

Long run forecasting of probable demand for a product of a company is generally for a period of 3
to 5 or 10 years.
1.Business planning It helps to plan expansion of the existing unit or a new production unit. Capital
budgeting of a firm is based on long run demand forecasting.
2.Financial planning: It helps to plan long run financial requirements and investment programs by
floating shares and debentures in the open market.
3.Manpower planning : It helps in preparing long term planning for imparting training to the
existing staff and recruit skilled and efficient labour force for its long run growth.
4. Business control : Effective control over total costs and revenues of a company helps to
determine the value and volume of business. This in its turn helps to estimate the total profits of
the firm. Thus it is possible to regulate business effectively to meet the challenges of the market. 5.
Determination of the growth rate of the firm : A steady and well conceived demand forecasting
determine the speed at which the company can grow.
6. Establishment of stability in the working of the firm : Fluctuations in production cause ups and
downs in business which retards smooth functioning of the firm. Demand forecasting reduces
production uncertainties and help in stabilizing the activities of the firm.
 7.Indicates interdependence of different industries : Demand forecasts of particular products
become the basis for demand forecasts of other related industries, e.g., demand forecast for
cotton textile industry supply information to the most likely demand for textile machinery, colour,
dye-stuff industry etc.,
8.More useful in case of developed nations: It is of great use in industrially advanced countries
where demand conditions fluctuate much more than supply conditions.

Levels Of Demand Forecasting

Demand forecasting may be undertaken at three different levels, viz., micro level or firm level,
industry level and macro level.
Micro level or firm level
This refers to the demand forecasting by the firm for its product. The management of a firm is
really interested in such forecasting. Generally speaking, demand forecasting refers to the
forecasting of demand of a firm.
Industry level
Demand forecasting for the product of an industry as a whole is generally undertaken by the trade
associations and the results are made available to the members. A member firm by using such data
and information may determine its market share.
Macro-level
Estimating industry demand for the economy as a whole will be based on macro-economic
variables like national income, national expenditure, consumption function, index of industrial
production, aggregate demand, aggregate supply etc, Generally, it is undertaken by national
institutes, govt. agencies etc. Such forecasts are helpful to the Government in determining the
volume of exports and imports, control of prices etc. The managerial economist has to take into
consideration the estimates of aggregate demand and also industry demand while making the
demand forecast for the product of a particular firm.
Criteria For Good Demand Forecasting
Apart from being technically efficient and economically ideal a good method of demand forecasting should
satisfy a few broad economic criteria. They are as follows:

Accuracy: Accuracy is the most important criterion of a demand forecast, even though cent percent accuracy
about the future demand cannot be assured. It is generally measured in terms of the past forecasts on the
present sales and by the number of times it is correct.

Plausibility: The techniques used and the assumptions made should be intelligible to the management. It is
essential for a correct interpretation of the results.

Simplicity: It should be simple, reasonable and consistent with the existing knowledge. A simple method is
always more comprehensive than the complicated one

Durability: Durability of demand forecast depends on the relationships of the variables considered and the
stability underlying such relationships, as for instance, the relation between price and demand, between
advertisement and sales, between the level of income and the volume of sales, and so on.
Flexibility: There should be scope for adjustments to meet the changing conditions. This imparts durability
to the technique.

Availability of data: Immediate availability of required data is of vital importance to business. It should be
made available on an up-to-date basis. There should be scope for making changes in the demand
relationships as they occur.

Economy: It should involve lesser costs as far as possible. Its costs must be compared against the benefits
of forecasts

Quickness: It should be capable of yielding quick and useful results. This helps the management to take
quick and effective decisions. Thus, an ideal forecasting method should be accurate, plausible, durable,
flexible, make the data available readily, economical and quick in yielding results.

Opinion poll method:

1) Consumers’ survey methods

2) Sales force opinion methods

3) Experts’ opinion method.

Statistical Methods:

1) Trends projection methods a) Fitting trend line by observation b) Least square liner regression c)
Time series analysis d) Moving average and annual difference e) Exponential Smoothing

Consumer Survey of Demand Forecasting

The goods and services are made available in the market for consumers. So who better than them
can understand the demand for the products? This is what this method is all about. Demand
forecasting through consumer survey could be done through three ways as mentioned below: •
Method of Total Enumeration • Survey through Sample • Method of End use Let’s see each of
these in detail: • Method of Total Enumeration: When the demand is forecasted on the basis of
possible quantities that all the consumers would purchase (in the period of forecast) then it is
called as the total enumeration method. In this method the consumers are questioned about their
future purchase amounts of product and responses of all of them are taken into consideration by
the forecaster. This method is the most straightforward one because the forecaster simply sums up
the possible demand of all the consumers to arrive at the future demand. The demand forecast
through this method is likely to be more exact because it is simply recording the data and arriving
at the forecast and is free from bias. But if the numbers of consumers are too many then this
method would not be viable. This method conjures a lot of time and is dearer. Also care needs to
be taken while recording the data as any error here would impact the final conclusion.

Survey through Sample: A sample is a part of the total. In this method the forecaster selects a
sample or panel of consumers who would generate sense of the demand for the product. To state
in other words the forecaster selects a group of consumers from the total consumers of the
product and then carries out the forecast considering the responses given by the sample group. In
this method the consumers may be asked questions pertaining to various factors that would impact
the demand for the product. These factors would be determinants of demand like the incomes of
the consumers, prices of related products, price of the commodity itself and so on. How is the
sample survey done? Direct interviews could be done or the forecaster may prepare detailed
questionnaire to generate responses from the consumers. The urgency of forecast and availability
of funds also would determine the kind of approach the forecaster would like to go with while
conducting sample survey. Once the sample survey is done the forecaster would then derive the
overall demand for the product on the basis of overall demand of the sample group. Sample survey
is quite useful in case of new products as it would give direct responses from the end users. The
changes in preferences of consumers, impact of promotionalactivities could also be understood
through sample survey. The forecaster need not carry out the survey for total consumers and
needs to focus on the sample group only. Though cost friendly and less time consuming this
method is subject to certain shortcomings like sampling error and lack of exact answers. The
cooperation by the consumers also plays an important role in arriving at the conclusions. Method
of end use: This is yet another method through which demand forecasting could be done. What if
the product has more than one use? It means a product may be used as an intermediate product
(used in the production of other products), used for final consumption and can also be exported as
well as imported. So how is demand forecasted through this method? Here the demand for the
intermediate product could be estimated. The demand of exports net of imports could be
estimated and demand of final consumption could be predicted. It means that here the forecast
will be done for three components (for the firms using intermediate products, for the exporters
and importers and survey of the end users). This all parts are then added up to arrive at the final
forecast. One of the distinguishing features of this kind of survey is that it provides forecast sector
wise because the forecaster gets an idea of demand for the product through various uses since it is
carried out by considering multiple uses of the product. The major drawback of this method lies in
the requisite of accurate production plan for the coming times by the firms.

2)Sales Force Opinion Methods

Under this method, the salesmen are nearest persons to the customers and are able to judge, their
minds and market. They better understand the reactions of the customers to the firms’ products
and their sales trends. The estimates of the different salesmen are collected and estimates sales
are predicted. These estimates are revised from time to time with changes in sales price, product,
designs, publicity programmes, and expected changes in competition, purchasing power, income
distribution, employment and population. It makes use of collective wisdom of salesmen,
departmental heads and top executives.
Advantages:
1. It is simple, common sense method involving no mathematical calculations.
2. It is based on the first-hand knowledge of salesman and the persons directly connected with
sales.
3. This method is particularly useful for sales of new product. It has the salesman’s judgment.
Disadvantages:
It is a subjective approach.
2. This method can be used only for short-term forecasting.
3. For long-term planning it is not useful.

3)Expert’s Opinions
Under this method expert’s opinions are sought from specialists in the field, from inside and
outside of the organizations or the organization collects opinions from such specialists; views of
expert’s published in the newspaper and journals, wholesalers and distributors for the company’s
products, agencies and professional experts. These opinions and views are analyzed and
deductions are made there from to arrive at the figure of demand forecasts.
Advantages:
1. Forecasts can be done easily and speedily.
2. It is based on expert’s views and opinions hence estimates are nearly accurate.
3. The method is suitable where past records of sales are not available.
4. The method is economical because survey is done through collection of the data. The expenses
of seeking the opinions and views of experts are much less than the expenses of actual survey.
Dis-advantages:
1. Estimates for a market segment cannot be possible.
2. The reliability of forecasting is always subjective because forecasting is not based on facts.
B)Barometric Methods
In barometric methods historical or cross sectional data are used to forecast the future probable
demand of a particular product by applying statistical models and mathematical, equations. These
methods are considered to be superior techniques of demand estimation. The important statistical
methods used in demand estimation are; A. Trend Projection Method In trend analysis past data
about the dependent and independent variables is used to project the sales in the coming years
assuming that factors responsible for the past trends will continue to behave in similar manner in
future also as they did in the past in determining the magnitude and trend of sales of a product. In
this method a data set of past sales are taken at specified time, generally at equal intervals to
depict the historical pattern under normal conditions. On the basis of derived historical pattern, the
 future sales of a company are project. The main aspect of this method lies in the use of time series
and changes in time series occur due to following reasons:-
1. Secular Trend: Secular Trend also known as long term trend indicate the gene ral tendency and
direction in which graph of a time series move in relatively over a long period of time. This can be
upward or downward trend, depending upon the behaviour.
2. Seasonal Trends: This trend reflects the changes in sales a company due to change in various
seasons or climates or due to festival season or sales clearance season etc. These changes are
repetitive in nature and related to twelve months period.
3. Cyclical Trends: These trends reflect the change in the demand for a product during diverse
phases of a business cycle i.e growth, boom, maturity, depression, revival, etc. 4. Random or
irregular trends: These changes arise randomly or irregularly due to unforeseen events such as
famines, earth quakes, floods, natural calamities, strikes, elections and crises. These changes take
place only in the short run and have their own impact on the sales of a firm. These trends cannot
be predicted. In trend projection method real problem is to separate and measure each of these
trends separately. In order to estimate the future demand of the product the impact of seasonal,
cyclical and irregular trends are eliminated from the data and only secular trend is used. The trend
in the time series can be eliminated by using any of the following method; I. Graphical Method, II.
The method of semi average, III. Moving average method and IV. The least square method I.
Graphical Method It is simplest method of trend projection. In this method periodical sales data is
plotted on a graph paper and a line is drawn through the plotted points. Then a free hand line is
drawn passing through as many points as possible. The direction of this free hand line or curve will
reflect the general trends whereas the actual trend line will show the seasonal, cyclical and
irregular trend.
Table-1
Sales data of XYZ Company
Years 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Sales(in 20 22 21 25 28 24 30 28 31 35 30
Lakhs)

Figure-1
Trend Projection Sales of XYZ Company

I. The Semi average method

In this method, first of all time series data of sale is divided into two equal parts and thereafter,
separate average sale is calculated for each half. The two values of averages are plotted on graph
corresponding to the time period. A straight line is then drawn by joining these two points. This line
become the trend line and is used to forecast future sale. It has been explained with following
example;
In the table -3, we have given time series sales data for 11 years (odd years) so in order to divide
the series into two equal parts the sale of the middle year (i.e 2010) has been eliminated for fitting
the trend. The average sale of first half i.e. 2005 to 2009 is ₹23.2 lakhs for the next half it is ₹30.8
lakhs. These two points have been plotted and joined with straight line to find the trend of the sale.

Year Sales (S) Average sale

2005 20 First half

2006 22 (20+22+21+25+28)/5

2007 21 =23.2

2008 25

2009 28

2010 24 ---

2011 30 (30+28+31+35+30)/5

2012 28 =30.8

2013 31

2014 35

2015 30

Figure-2
 Trend Projection Sales of XYZ Company

II. Moving average Method

Moving average method is very widely used in practice. Under this method, moving average is
calculated. One has to decide what moving year average – 3year or 5year or 7year should be
taken up and it depends upon the periodicity of the data. It is determined by plotting the data on
the graph paper and noticing the average time interval of successive peaks or trough. After
deciding the moving year average, moving average of the given sales data is calculated and these
averages are plotted on the graph paper to fit the trend. It has been explained with help of
following example;

Table-3
Sales data of XYZ Company (₹, lakhs)

Year Sales (S) 3 years moving

Average

2005 20 -

2006 22 (20+22+21)/3 =21.00

2007 21 (22+21+25)/3 =22.67

2008 25 (21+25+28)/3 =24.67

2009 28 (25+28+24)/3 =25.67

2010 24 (28+24+30)/3 =27.33

2011 30 (24+30+28)/3 =27.33

2012 28 (30+28+31)/3 =29.33

 2013 31 (28+31+35)/3 =31.33

2014 35 (31+35+30)/3 =32.00

2015 30 -

3 Year Moving Average Trend Projection of

Sales
40

30

20
Year 2006 2007 2008 2009 2010 2011 2012 2013

Actaul Sales 3 years moving average

10
Figure-3

III. The least square method

0
Fitting trend equation or popularly known as least square method is a scientific, formal and popular
method of projecting the trend line. In this method a trend line is fitted with the help of straight line
regression equation i.e
S = a + bT
where S = annual sales, T = Time, a and b are constants. The coefficients a and b are calculated by
solving following two equations;
i) ΣS = Na + ΣT
ii) ΣST = a ΣT +
2
bΣT Where,
IV. ΣS = Sum of the original sales of N
years (S)N= Number of years
V. ΣT = Sum of deviations of the years taken from a central
periodΣT2 = Sum of the squared deviations of T values
VI. ΣST = Sum of the product of the deviation and corresponding sale
VII. Taking the data given in table-4 given below, the regression equation i) and ii) are given
below;i) 294 = 11a + 0
VIII. ii) 148 = a*0 +110b by solving these two equation we get; S
= 28.73 + 1.345T
IX. With the help of this equation, it is quite easy to forecast the sale for any future year i.e
 for the years 2018, 2020 or 2022 by taking T as the deviation from the base year (2010)
and t in this casewill be 8th, 10th, and 12th year. It can be calculates as follow;
X. 2018 S2018 = 28.73 + 1.345(8) = ₹39.50 lakhs
XI. 2020 S2020 = 28.73 + 1.345(10) = ₹42.18 lakhs
XII. 2022 S2022 = 28.73 + 1.345(12) = ₹44.87 lakhs
XIII. In order to fit the trend line the computed sales value (S c) are to be plotted on the graph
paper asgiven in the figure-4

Year Sales T T2 ST Sc (Computed Sales value)

(S)
200 20 -5 25 -100 28.73 + 1.345(-5) =22.01
5
200 22 -4 16 -88 28.73 + 1.345(-4) = 23.35
6
200 21 -3 09 -63 28.73 + 1.345(-3) = 24.70
7
200 25 -2 04 -50 28.73 + 1.345(-2) = 26.04
8
200 28 -1 01 -28 28.73 + 1.345(-1) = 27.39
9
201 24 0 00 0 28.73 + 1.345(0) = 28.73
0
201 30 1 01 30 28.73 + 1.345(1) = 30.08
1
201 28 2 04 56 28.73 + 1.345(2) = 31.42
2
201 31 3 09 93 28.73 + 1.345(3) = 32.77
3
201 35 4 16 140 28.73 + 1.345(4) = 34.11
4
201 30 5 25 150 28.73 + 1.345(5) = 35.46
5
ΣS= 294 ΣT=0 ΣT2=110 ΣST=148

The least square method is very popular method used in demand forecasting because it is very
easy and less expensive method.

Figure-4
 Trend Projection Sales of XYZ Company

Trend Line using least Sequre Method

40

35

30

25

20

15
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
10
Years
5

0
Demand Forecasting For a New Product

Demand forecasting in case of new products is not easy as in case of established products. In case
new product the firms will not have any past experience or past data for this purpose. It requires
an intensive research of the economic and competitive features of the product should be made
to make efficient forecasts.

Professor Joel Dean, however, has suggested a few guidelines to make forecasting of demand for
new products.

a. Evolutionary approach

The demand for the new product may be considered as an outgrowth of an existing product. For
e.g., Demand for new Tata Indica, which is a modified version of Old Indica can most effectively be
projected based on the sales of the old Indica, the demand for new Pulsor can be forecasted based
on the sales of the old Pulsor. Thus when a new product is evolved from the old product, the
demand conditions of the old product can be taken as a basis for forecasting the demand for the new
product.

b. Substitute approach

If the new product developed serves as substitute for the existing product, the demand for the new
product may be worked out on the basis of a ‘market share’. The growths of demand for all the
products have to be worked out on the basis of intelligent forecasts f or independent variables that
influence the demand for the substitutes. After that, a portion of the market can be sliced out for the
 new product. For e.g., A moped as a substitute for a scooter, a cell phone as a substitute for a land
line.In some cases price plays an important role in shaping future demand for the product.

Delphi method
The Delphi method is a type of forecasting model that involves a small group of relevant experts
who express their judgment and opinion on a given problem or situation. The expert opinions are
then combined with market orientation to come up with results and develop an accurate
forecast.
The Delphi method is performed in such a way that each expert is questioned individually to gather
their insights. This helps to prevent bias and ensures that the company’s forecast is based solely on
their own expert opinion.
Furthermore, other employees or outsourced parties collect, summarize, and analyze experts’
responses. They may pose additional questions to the participants who can then reconsider their
original responses in order to come up to a meeting point or final consensus that would be
beneficial to the company.
Highlights of Delphi method
Some of the highlights of this model include:
 Freedom of expression — Since each expert is questioned individually, they have the freedom to
express their own opinion without feeling peer pressure,
 Ability to make up their mind and give a second thought — The experts can change their opinion
and provide additional information in case they have reassessed the problem,
 Consistent feedback — After each round, the participants are informed about the opinions of other
group members and then make discussions, and
 Quantitative results — This type of model is qualitative in nature, however there is possibility to
analyze the results quantitatively.
When to use Delphi?
The Delphi method can be used to:
 Predict trends in sales,
 Forecast outcomes in economic development,
 Identify risks and opportunities,
 Create work breakdown structures, and
 Compile a report from opinions.
Other situations where the Delphi method makes sense is when you:
 Want to gather subjective statements from a larger group,
 Find it difficult to perform a face-to-face discussion due to the group size,
 Need to retain the anonymity of the participants, and
 Feel there is a dominant person in the group who could interfere with the discussion.

Market research model

Market research is a qualitative forecasting model that evaluates the performance of a business’s
products and services by interviewing potential customers about them. Their reactions and
responses are recorded, and then they’re analyzed in order to come up with a sales forecast.
 This model can be performed by staff members or third-party agencies (specialized in market
research) by:
 Telephone,
 Opinion poll,
 Personal interviews, or
 Questionnaires.
Some examples of market research strategies include:
 Focus groups,
 Consumer surveys, or
 Product testing.
However, the strategies might be adapted based on the current market conditions and
challenges.
These techniques are used to gather valuable insights from consumers so that the company
understands which products or services to continue launching and which ones need to be
revised.
Highlights of market research
The market research model can help companies:
 Create consumer-oriented marketing policies,
 Study marketing problems in order to come up with a solution,
 Minimize the gap between consumers and manufacturers,
 Introduce new products,
 Identify potential markets, and
 Choose marketing methodologies.
When to use market research?
The market research model can be used:
 Before the launch of a new product or service,
 Once a product or service has been launched in order to find out if it’s been accepted, or
 As a continuous plan that will help businesses get information about its market standing.
Conducting market research is also beneficial when you need to:
 Estimate the market size,
 Define potential customers,
 Find out why sales are declining, or
 Support business growth.

Associative Forecasting Methods: Regression and Correlation Analysis

Unlike time-series forecasting, associative forecasting models usually consider
several variables that are related to the quantity being predicted. Once these related variables
have been found, a statistical model is built and used to forecast the item of interest. This
approach is more powerful than the time-series methods that use only the historical values for
the forecast variable.
Using Regression Analysis for Forecasting
We can use the same mathematical model that we employed in the least-squares method of
trend projection to perform a linear-regression analysis. The dependent variables that we want
to forecast will still be yn . But now the independent variable, x , need no longer be time.
Standard Error of the Estimate
The forecast of $3,250,000 for Nodel’s sales in Example 12 is called a point estimate of y .
The point estimate is really the mean , or expected value , of a distribution of possible value
of sales.
Correlation Coefficients for Regression Lines
The regression equation is one way of expressing the nature of the relationship between two
variables. Regression lines are not “cause-and-effect” relationships. They merely describe the
relationships among variables. The regression equation shows how one variable relates to the
value and changes in another variable.
Multiple-Regression Analysis
Multiple regression is a practical extension of the simple regression model we just explored.
It allows us to build a model with several independent variables instead of just one variable.
Monitoring and Controlling Forecasts
Once a forecast has been completed, it should not be forgotten. No manager wants to be
reminded that his or her forecast is horribly inaccurate, but a firm needs to determine why
actual demand (or whatever variable is being examined) differed significantly from that
projected. If the forecaster is accurate, that individual usually makes sure that everyone is
aware of his or her talents.
Adaptive Smoothing Adaptive forecasting refers to computer monitoring of tracking signals
and self-adjustment if a signal passes a preset limit.
Focus Forecasting Rather than adapt by choosing a smoothing constant, computers allow us
to try a variety of forecasting models. Such an approach is called focus forecasting. Focus
forecasting is based on two principles:
1. Sophisticated forecasting models are not always better than simple ones.
2. There is no single technique that should be used for all products or services.
Forecasting in the Service Sector
Forecasting in the service sector presents some unusual challenges. A major technique in the
retail sector is tracking demand by maintaining good short-term records. For instance, a
barbershop catering to men expects peak flows on Fridays and Saturdays.
Specialty Retail Shops Specialty retail facilities, such as flower shops, may have other
unusual demand patterns, and those patterns will differ depending on the holiday. When
Valentine’s Day falls on a weekend, for example, flowers can’t be delivered to offices, and
those romantically inclined are likely to celebrate with outings rather than flowers.
Fast-Food Restaurants Fast-food restaurants are well aware not only of weekly, daily, and
hourly but even 15-minute variations in demands that influence sales. Therefore, detailed
forecasts of demand are needed .

Web-based forecasting is a method of using the internet to predict future events, trends, or
outcomes based on historical data and statistical analysis. Web-based forecasting systems can use
a variety of techniques to improve accuracy and efficiency, such as:
 Machine learning
 Econometric modeling
 Regression analysis
 Hybrid neural network
 Fuzzy logic technique
 Traditional statistical methods
Web-based forecasting systems can offer several benefits, including: Removing geographical
barriers, Reducing information dissemination costs, Enabling collaboration, and Preventing the
influence of leading personalities.
Some examples of web-based forecasting systems include:
 Web-Based Intelligent Forecasting System (WIFS): A prototype software that uses a hybrid neural
network, fuzzy logic technique, and traditional statistical methods
 Online Forecasting Calculator Tool: A tool from Call Centre Helper that can be used to make
forecasts

Collaborative Planning, Forecasting, and Replenishment (CPFR) is a web-based tool that helps
supply chain trading partners coordinate their planning and execution of customer demand. CPFR is
a strategic approach that involves collaboration between trading partners, such as manufacturers,
retailers, and suppliers, to improve supply chain efficiency and agility.
CPFR helps to:
 Improve supply chain efficiency
CPFR can reduce inventory costs and enhance the performance of the entire supply chain.
 Optimize sales forecasts
CPFR can help to optimize sales forecasts and respond to changes in market demand.
 Reduce silos
CPFR can streamline supply chain planning activities and multiple S&OP processes.
CPFR involves trading partners sharing information and jointly planning key supply chain activities,
such as: Business planning, Sales forecasting, and Replenishment of finished goods and raw
materials.

The Collaborative Planning, Forecasting, and Replenishment (CPFR) process is a business concept
that involves trading partners working together to improve supply chain efficiency:
The CPFR process typically includes the following steps:
 Planning
Trading partners work together to create a unified sales and production plan. This plan is based on
real-time data, historical information, and business analytics.
 Forecasting
Trading partners work together to create collaborative sales forecasts. These forecasts take into
account historical sales data, market trends, and external factors.
 Inventory management
Trading partners work together to monitor and manage inventory levels. This helps to coordinate
deliveries and minimize stockouts and overstock situations.
 Other steps in the CPFR process include:
 Developing an agreement on how communication will be carried out
 Creating a joint business plan
 Identifying exceptions and deviations in the sales forecast
 Resolving exception items
 Creating an order forecast
 Identifying exceptions and deviations in the order forecast
 Generating orders
CPFR Process Model Figure 8-1: CPFR Model (After J.D.Edwards White paper, 2003) The model as
shown in Figure 8-1 is a simplified version of the 9-step model. The model comprises:

The CPFR model enables significant scope and depth of collaboration across supply chains. CPFR
involves a number of business processes integrated between a number of supply chain partners eg
between a retailer and a supermarket. There is usually a few lead partners who select those
processes where CPFR is adopted. There is data exchange between the partners and include
suppliers taking responsibility for replenishment on behalf of their customer. Synchronized
forecasting is also involved in CPFR. The individual information systems are coordinated for
planning and replenishment pruposes. From the data of actual consumer demand extracted from
POS, product development, marketing plans, production planning and transport planning are
seamlessly integrated with forecasts.
Planning Phase In the planning phase, there are 2 stages:
Front-end Agreement The parties involved establish the guidelines and rules for the collaborative
relationship. The agreement includes a common basis for co-operation, trust and availability of
competency resources. All parties are bound to make available and ready these competencies and
resources for the system to work. The Business Intelligence modules allow partners to define and
measure specific KPIs. The agreement also includes mechanism to handle disagreements and
differences.

Joint Business Plan

The parties involved create a business plan that takes into account their individual corporate
strategies and defined category roles, objectives and tactics. This includes product types, minimum
order quantities, lead time and order frequency. The business plan becomes the communication
tool among the supply chain partners. The front-end agreement should produce a long-term pact
spanning the life of the business. Obviously, an enormous amount of information will flow between
partners. These are: Who should get what? When? Where? How much should they get?

Forecasting Phase
The stages are: Sales-forecast Retailer point-of-sales data, causal information and information on
planned events are used by one party to create an initial sales forecast. This forecast is then
communicated to the other party and used as a baseline for the creation of an order forecast.

Identify exceptions for sales forecast Items that fall outside the sales forecast constraints set in the
front-end agreement are identified. The criteria for exceptions are stated in the Front-end
Agreement. Examples of such items are seasonal products.
Resolve / collaborate on exception items Exceptions are easily identified and messages are sent to
reconcile unusual items. Each contributor (partner, supplier, and customer) becomes an integral
part of the real-time collaborative process. The final enterprise forecast is the combination of the
most accurate and timely information available. The parties negotiate and produce an adjusted
forecast.
Create order forecast The order forecast relies on point-of-sale (POS) data, causal information,
and inventory strategies to generate a specific forecast that supports the shared sales forecast.
Identify exceptions for order forecast Items that fall outside the order forecast constraints set
jointly by the parties involved are identified.
Resolve / collaborate on exception items The parties negotiate (if necessary) to produce an
adjusted order forecast.

Replenishment Phase
The single stage is: Order Generation The final step in the CPFR process is generating the order
and promising the delivery. The order forecast is translated into a firm order by one of the parties
involved.

The essence of maintaining positive relationships with partners and customers is to deliver on
promises.
5. Benefits of CPFR Improved customer service trough better forecasting techniques More
reliable forecasting allows a more effective way to anticipate consumer demand across the entire
 supply chain and therefore allow the business to plan production capacity accordingly. Risks for
stock-outs is reduced which improves customer fulfillment orders which thereby increases
revenue, delivery and improved customer service. Lower Inventories for higher profits
 Accurate predictions of demand as mentioned before will reduce stock-outs and provide a more
efficient understanding of production needs. Safety stock inventory for over production would be
reduced which decreases carrying costs, storage space and potential spoilage/obsolescence.
Additionally, there is improved material flow and release of working capital that can be used in
other areas of the production instead of being tied up in inventory.
Improved ROI on Technology investment Effective CPFR technology solutions benefit both
manufacturers and retailers from reduced overhead costs because several inefficiencies are
eliminated, i.e., antiquated manual processes, custom integrations of different partner IT systems
and information searching of multiple sources/systems. Improved relationships between trading
partners
 Develop when collaboration takes place. Trading partners gain a better understanding of
respective businesses by regularly exchanging information and establishing direct communication
on channels and create a win-win situation.
Cost reduction Will occur when production schedule and agreed forecasts are aligned. Costs are
reduced by decreasing set-up times, effort duplication and variations. There is also efficient
production capacity utilization since planning information is more reliable.

Cons of CPFR
 Implementation Complexity:
CPFR implementation can be complex and resource-intensive. Adopting collaborative planning
requires not only the integration of technology but also changes in organizational processes and
culture. This complexity can pose challenges, particularly for businesses with existing systems that
may not easily align with CPFR principles.
 Data Accuracy and Trust Issues:
Reliance on accurate and timely data is vital for the success of CPFR. Issues such as data
inaccuracies, discrepancies, or a lack of trust between trading partners can undermine the
effectiveness of collaborative planning. If one party doubts the accuracy of shared information, it
can lead to suboptimal decision-making and collaboration breakdowns.
 Dependency on Partner Cooperation:
CPFR relies on the active cooperation and engagement of all trading partners involved in the supply
chain. If one partner fails to actively participate or share relevant information, it can disrupt the
collaborative planning process. Achieving a synchronized and effective supply chain requires
commitment and coordination from all parties, and any lack of alignment can hinder the desired
outcomes.
 What is Forecast Accuracy?
Forecast accuracy is the means of measuring how well a demand forecast has predicted actual
outcomes or values of sales. Additionally, forecast accuracy ascertains the reliability and
effectiveness of one’s forecasting models and techniques currently in-use. Determining one’s
forecast accuracy provides more insight into a company’s market, allowing for more effective
decision-making with regard to stock numbers and levels. The future benefit of ensuring that the
right amount of inventory is available at the right time helps mitigate any issues that could cause
major revenue losses such as over or understocking.
Forecast accuracy is also commonly used across various fields, such as sales forecasting, demand
planning, and financial forecasting. In sales forecasting, accurate predictions help organisations
estimate future sales volumes, optimise inventory levels, and streamline production processes. Like
sales forecasting, demand planning places its focus on stock levels, ensuring the availability of
products to meet customer demands without running into excess inventory or stock-out issues.
Financial forecasting utilises forecast accuracy to make informed monetary decisions: budgeting,
resource allocation, and investment strategies.

Why do we Measure Forecast Accuracy?

By comparing the predicted values with the actual outcomes, forecast accuracy provides insights
into the quality of the forecasts and the level of confidence that can be placed in them. By regularly
assessing and improving forecast accuracy, businesses can enhance their planning processes,
optimise resource allocation, and make more informed and strategic decisions.
1. Reduced Costs: Through avoiding excessive stockpiling or shortage situations, forecast accuracy
allows for better monetary management and cost savings. With reliable forecasts and aligning their
production and procurement activities with a more accurate expected demand, forecast accuracy
helps companies minimise holding costs, obsolescence, and carrying costs associated with over and
under-stockings, resulting in reduced costs.
2. Improved customer satisfaction: Accurate forecasts play a crucial role in ensuring product
availability and meeting customer demands. Accurate predictions of customer needs and demands
allow for better resource management, which helps in avoiding stock-outs, and betters planning of
production and supply chain activities. Customers are taught to rely on receiving their products in a
timely manner, fostering positive experiences and long-term loyalty.
3. Effective inventory management, resource allocation, and production planning: Forecast accuracy
directly impacts inventory management by providing insights into the expected demand for
products or services. Accurate forecasts help in efficient resource allocation, by aligning
production, labour, and procurement activities with the projected demand. The predicted
inventory levels furthermore dictate how each resource will be used and in which quantities.
Additionally, precise forecasts aid in effective production planning, allowing businesses to
streamline their operations, avoid bottlenecks, and meet customer demands efficiently.
How is Forecasting Accuracy Measured?
Now that we know why we measure forecasting accuracy, let us look at how it is measured in three
different ways.
Mean Absolute Error (MAE): A measure of forecast accuracy that calculates the average absolute
difference between predicted values and the actual values. This metric assesses how close the
forecast has come to the actual outcomes, regardless of the direction of the errors. It is a
straightforward interpretation of the average error magnitude. MAE is calculated by summing up
the absolute difference of your forecast and the actual value, divided by your sample size.
MAE = 1/n * Σ |Actual - Forecast|
Mean Squared Error (MSE): MSE is widely used in forecasting as it gives more weight to outliers
and larger errors, making it useful when significant errors should be penalised more heavily.
Mean Absolute Percentage Error (MAPE): MAPE expresses forecast errors as a percentage of the
actual values. It provides an assessment of the average magnitude of the forecast errors in relation
to the scale of the actual values. MAPE is commonly used in situations where percentage accuracy
is more meaningful (such as demand forecasting). It is calculated by taking the absolute difference
between the forecast and actual values, and dividing this difference by the actual value. To get a
percentage form, it is important to multiply this division by 100.
MAPE = 100/n * Σ|Actual - Forecast|/Actual

Short answer questions

i)What is Demand Management?Explain in concise form
ii)What do you understand the needs of Demand Forecasting?
iii)Name the various methods of demand forecasting and define.
iv) What is the delphi method? Describe its main advantages and
limitations.

Long Answer Questions

i)What is Demand Forecasting.
ii)Explain the key objectives and significance of Demand Forecasting?
iii)Give the meaning of Web Based Forecasting and explain in detail CPFR.

Case Study
Given the following data:
Period Number of Complaints
1 60
2 65
3 55
4 58
5 64
Prepare a forecast for period 6 using each of these approaches

 A three-period moving average