DEMAND FORECASTING & COLLABORATIVE

PLANNING, FORECASTING, &

REPLENISHMENT
 Learning Objectives
You should be able to:
–Explain the role of demand forecasting in a
supply chain.
–Identify the components of a forecast
–Compare and contrast qualitative and
quantitative forecasting techniques
–Assess the accuracy of forecasts
–Explain collaborative planning, forecasting, and
replenishment
2
 Chapter Five Outline
• Introduction
• Matching Supply and Demand
• Forecasting Techniques
–Qualitative Methods
–Quantitative Methods
• Forecast Accuracy
• Collaborative Planning, Forecasting, and
Replenishment
• Software Solutions
3
 Introduction
• Competitive environment – more effective demand-
driven supply chain to respond quickly to market
changes. (Customers, Competitors, Seasonal etc.)
• Used to be “push” market environment. Now it’s “pull”
market environment.
• Matching supply and demand as closely as possible.
• Forecasting - an estimate of future demand
• The goal is to minimize forecast error.
• Have to consider the factors that influence demand
• Improved forecasts benefit all trading partners in the
supply chain.
4
 Demand Forecasting
• Estimate of future demands – planning and
business decisions
• Future are unknown – errors.
• Choice of appropriate forecasting techniques –
to reduce errors.
• To consider factors that influence demand,
impact of these factors, whether it still influence
future demand

5
 Benefits of Good Forecast
• Allows the right amount of products (Raw
materials, components and MRO)
• Produce the right types and amount of
products
• To deliver the right number of products at
the right time.

6
 Forecasting Techniques

Qualitative forecasting is based on opinion and

intuition.

Quantitative forecasting uses mathematical

models and historical data to make forecasts.

7
 Forecasting Techniques- Cont.
Qualitative Forecasting Methods

Generally used when data are limited, unavailable, or not

currently relevant. Forecast depends on skill & experience
of forecaster(s) & available information.

Four qualitative models used are:

1. Jury of executive opinion
2. Delphi method
3. Sales force composite
4. Consumer survey
8
 Aim/Results of Accurate Forecast
• Lower inventories
• Avoid or Minimize stock-outs
• Smoother production plans
• Reduced costs, and
• Better customer service.

9
 Jury Opinion
• Group of senior management executives.
• Knowledgeable about the market, competitors
and business environment.
• Collectively develop the forecast.
• Long range plan
• New product development

10
 Delphi Method
• Separate interviews on a group of internal and external
experts on future events and long term demand forecast.
• Answers accumulated, summarized and sent to the
experts for the next round consideration. Changes are
allowed
• This process will go on until a consensus reached.
• Good for high-risk technology forecasting, expensive
projects, major and/or new product introductions.

11
 Sales Force Composite
• Sales are the closest to the market.
• Generated based on sales force
knowledge in the market
• Reliable but subject to individual biases.

12
 Consumer Survey
• Questionnaires
• To determine buying habits, new product
ideas, opinion on existing products

13
 Forecasting Techniques- Cont.
Quantitative Methods
Time series forecasting- based on the assumption that
the future is an extension of the past. Historical data is
used to predict future demand. Use a series of
observation in chronological order to develop forecasts.

Associative forecasting- assumes that one or more

factors (independent variables) predict future demand.

It is generally recommended to use a combination of

quantitative and qualitative techniques.
14
 Forecasting Techniques- Cont.
Components of Time Series- Data should be
plotted to detect for the following components:
Trend variations: Long term movement up or down in
a time series
Seasonal variations: show peaks and valleys that
repeat over a consistent interval such as hours, days,
weeks, months, years, or seasons
Random variations: unpredictable movement from one
time period to the next

15
 Time series with randomness

Figure 5.1
 Time series with
Trend and Seasonality

Figure 5.2
 Forecasting Techniques- Cont.
Time Series Forecasting Models
– Simple Moving Average Forecasting Model.
Simple moving average forecasting method uses
historical data to generate a forecast. Works well
when demand is fairly stable over time.

18
 Simple Moving Average Forecast
Period
1
Demand
1600
• Using the data provided,
2 2200 calculate the forecast for
3 2000 period 5 using
4 1600
5 2500
• four-period simple moving
6 3500 average
7 3300
8 3200
• Forecast for period 5 =
9 3900 (1600 + 2200 + 2000 +
10 4700
11 4300
1600)/4=1850
12 4400
19
4
Using Excel Spreadsheet

20
 Weighted Moving Average Model

• Weighted Moving Average Model – A form

of the moving average model that allows
the actual weights applied to past
observations to differ.
 Example
Period Demand Based on data provided, calculate the forecast for
1 1600 period 5 using a four-period weighted moving
2 2200 average.
3 2000 The weight of 0.4, 0.3, 0.2, 0.1 are assigned to the
4 1600
most recent, second most recent, third most recent
and most fourth recent respectively.
5 2500
Forecast for period 5 =
6 3500
0.1(1600) + 0.2(2200) + 0.3(2000) + 0.4(1600)
7 3300
= 1840
8 3200

9 3900

10 4700

11 4300

12 4400

22
Using Excel Spreadsheet

23
 Causal Forecasting Models
 Linear Regression
 Multiple Regression
 Examples:
 Linear Trend Forecast
• Forecast estimated using simple linear regression to fit a
line to a series of data occurring over time – simple trend
model.
• The trend line is determined using least square method.
Ŷ = b0 + b1x
– where
Ŷ = forecast or dependent variable
bₒ = intercept of the line
b1 = slope of the line
x = time variable

25
b0 = intercept of the line
b1 = slope of the line

b0 = Σy – b₁ Σx/n
b1 = n Σ(xy) - Σx Σy/n Σx² – (Σx)²

26
 Forecasting Techniques- Cont.
Associative Forecasting Models- One or several external variables
are identified that are related to demand
– Simple regression. Only one explanatory variable is used and is
similar to the previous trend model. The difference is that the x
variable is no longer a time but an explanatory variable.
Ŷ = b0 + b1 x
– where
Ŷ = forecast or dependent variable
x = time (period) or independent variable
b0 = intercept of the line
b1 = slope of the line

27
 Linear Trend Forecasting
The demand for toys produced by XXXX Co. as shown below
Period (x) Demand
1 1600

2 2200

3 2000
4 1600
5 2500

6 3500

7 3300

8 3200
9 3900

10 4700

11 4300
12 4400
Total (Σ) 78 37200

28
 Period (x) Demand x² xy b0 = Σy – b₁ Σx/n =
1 1600 1 1600

2 2200 4 4400 b1 = n Σ(xy) - Σx Σy/n Σx² – (Σx)²

3 2000 9 6000 b1 = 12(28200) – 78(37200)/12(650) -78²
4 1600 16 6400 = 286.71
5 2500 25 12500 b0 = 37200 -286.71(78)/12 =1236.4
6 3500 36 21000 The trend line is
7 3300 49 23100
Ŷ = b0 + b1x
8 3200 64 25600
Ŷ = 1236.4 + 286.7(x)
9 3900 81 35100
to forecast demand for period 13
10 4700 100 47000
Ŷ = 1236.4 + 286.7(13) = 4964 toys
11 4300 121 47300
12 4400 144 52800
l (Σ) 78 37200 650 282800

29
Forecasting Techniques- Cont.
– Multiple regression. Where several explanatory variables are used to
make the forecast.

Ŷ = b0 + b1x1 + b2x2 + . . . bkxk

– where
Ŷ = forecast or dependent variable
xk = kth explanatory or independent variable
b0 = intercept of the line
bk = regression coefficient of the independent variable xk

30
 Forecast Accuracy
The formula for forecast error, defined as the difference between
actual quantity and the forecast, follows:
Forecast error, et = At - Ft
where
et = forecast error for Period t
At = actual demand for Period t
Ft = forecast for Period t
Several measures of forecasting accuracy follow:
– Mean absolute deviation (MAD)- a MAD of 0 indicates the
forecast exactly predicted demand.
– Mean absolute percentage error (MAPE)- provides
prerspective of the true magnitude of the forecast error.
– Mean squared error (MSE)- analogous to variance, large
forecast errors are heavily penalized

31
 Forecast Accuracy

How do we know:
• If a forecast model is “best”?
• If a forecast model is still working?
• What types of errors a particular forecasting
model is prone to make?

Need measures of forecast accuracy

 Forecast Accuracy Measure
Period (x) Demand Forecast Error (e) Absolute e² Absolute
Error Error %
1 1600 1523 77 77 5929 4.8

2 2200 1810 390 390 152,100 17.7

3 2000 2097 -97 97 9409 4.9

4 1600 2383 -783 783 613,089 48.9

5 2500 2670 -170 170 28.900 6.8

6 3500 2957 543 543 294,849 15.5

7 3300 3243 57 57 3249 1.7

8 3200 3530 -330 330 108,900 10.3

9 3900 3817 83 83 6889 2.1

10 4700 4103 597 597 356,409 12.7

11 4300 4390 -90 90 8100 2.1

12 4400 4677 -277 277 76,729 6.3

Total 0 3494 1,664,552 133.9
Average RSFE 291.17 138,712.7 11.16
MSE
MAD
MAPE
33