Demand Forecasting

McGraw-Hill/Irwin
Copyright 2008 by The McGraw-Hill Companies, Inc. All rights reserved.

Demand Management
Processing,
Influencing, &
Anticipating
Demand
Buy
Sell
Buy
Store
Move
Make
Sell
Store
Move
Sell
Make
Make
Move
Move
Buy
4-2
Managing the sell side of a business
Plant
Plant
Plant
Warehouse
S
u
p
p
l
i
e
r
s

C
u
s
t
o
m
e
r
s

Supply-Demand Management
"Make, Move, Store"
Supplier
Relationship
Management
"Buy"
Customer
Relationship
Management
"Sell"
4-3
Key questions
1. What is the scope of demand management?
2. What does order processing involve; why is it an important area for management
attention?
3. What is customer profit potential, & how is it relevant for influencing demand?
4. What are 5 alternatives for improving forecast accuracy, what do they mean, & how
can they be applied?
5. How do the tactics of part standardization & postponement of form or place help
improve forecast accuracy?
6. What is the difference between long term & short term forecasting?
7. What are 4 long term forecasting methods; what are the risks of
salesperson/customer input?
8. What are the components of demand, & which component is not forecasted?
9. How do the moving average, Winters, & focus forecasting methods work?
10. What is the role of the number of periods in the moving average method, & the
smoothing parameters in the Winters method?
11. What is the purpose of filtering, & why is it important for computer-based
forecasting?
12. What do the following principles of nature mean & how are they relevant for
demand management? (1) law of large numbers, (2) trumpet of doom, (3) recency
effect, (4) hockey stick effect, (5) Pareto phenomenon
13. What are the managerial insights from the chapter?

4-4
Road map
Processing Demand

Influencing Demand

How to Improve Forecast Accuracy

Long Term Forecasting

Short Term Forecasting

Summary
4-5
Scope of demand management
So what is demand management?

Concerned with processing, influencing,
and anticipating demand

Well begin with processing demand or,
in more common terms, order
processing or order fulfillment
4-6
Order processing
Order processing is usually viewed to span
order booking to order shipment

Example steps?

Customer validation, order entry, credit checking,
pricing, design changes, availability checks, delivery
time estimation, notification of shipment, notification
of delays
Processing Demand
4-7
Processing Demand
CUSTOMER

ORDER ENTRY AND
CHECKING
Customer Validation
Credit Control Operations

ER
P
INVOICING
SHIPPING
CUSTOMER SERVICE
ORDER
INTERRUPTION

ORDER
PICKING AND
ASSEMBLY
RETURNS

4-8
Characteristics
Can be a complex & time consuming process
dealing largely with information flow

Susceptible to ad hoc modifications over time in
response to problems (e.g., extra credit check added
due to expensive nonpaying customer a few years
ago)

A major customer contact point with
organization

Can significantly impact customer perceptions

IT advances & high customer impact

A potential profitable target for improvement
Processing Demand
4-9
Example 1
Benetton

Electronic loop linking sales agent, factory, & warehouse

If not available, measurements transferred to knitting
machine for production

Benetton uses a single warehouse

Staffed by 8 people & about 230,000 pieces shipped daily
Processing Demand
4-10
Example 2
K-Mart and MasterLock

Policy for mistake in shipment or invoice

Strike 1: $10,000, Strike 2: $50,000, Strike 3: lose
business

MasterLock revamped their order processing function
Processing Demand
4-11
Example 3 customer tools
Processing Demand
Amazon online order tracking
4-12
Example 4 customer tools
Processing Demand
UPS online order tracking
4-13
Example 4 continued
Processing Demand
UPS online tools
4-14
Road map
Processing Demand

Influencing Demand




Summary
4-15
Measure customer profit potential
Some customers are more profitable than others

Advancing technologies more practical to estimate profit
potential of individual customers

Can guide efforts/investments for customer retention &
acquisition . . . investments to influence demand

E.g.,

Electronics manufacturer: reviews historical customer profit before
sending service contract renewal

Wireless phone firm: churn scores & lifetime value estimates
influence # of customer contacts & attractiveness of offerings

Ongoing development of data mining methods
Influencing Demand
A simple idea
4-16
Road map
Processing Demand

Influencing Demand




Summary
4-17
Motivating example 1
Sunbeam

Improved forecasting led to 45% reduction in
inventory

Included estimates from top 200 customers
Forecasting Alternatives
4-18
Apple

A history of problems forecasting demand

Many components sourced from 1 supplier -
accurate forecasts are critical

Over $1 billion in unfilled orders during the
crucial holiday season. The CEO (Spindler)
ousted a few months later
4-19
IBM

Badly misjudged demand in PC business in 1996
went from being profitable in 1995 to a $200
million loss through 1
st
half of 1996
4-20
Christmas 1999 & e-commerce takes
off

Large unanticipated increase in Internet orders
didnt ship on time

E.g., Many Toys R Us Christmas orders not delivered
until March I will never buy online again
4-21
Improvement alternatives
Change the forecasting method

Collect more or different data

Analyze the information differently

E.g., involve more people, new forecasting software, spend more
time manually reviewing, focus groups etc.

Change operations or operating policies

Introduce early warning mechanisms

Take advantage of the law of large numbers

Reduce information delays & leadtimes (trumpet of doom)

Reduce demand volatility
4-22
Early warning
Change policies so that some (or more)
customers provide earlier commitment of
future demand, e.g.,

Early bird program for builder markets discount for
60-day advance order

Invite large buyers to Aspen in February to view next
years skiwear line, & encourage orders

Commitment = asking customers how much
they are likely to buy next quarter
4-23
Law of large numbers
As volume increases, relative variability decreases

Postponement in form or place, e.g.,

Dell configure your own PC

From full product line at 12 regional DCs to full product line at a
single super DC, with 10% of product line stocked at 11 regional
DCs (i.e., fast movers that account for 70% of sales)

Part standardization, e.g.,

Arbys sandwich wrappers; plastic lids with push down drink
indicator

Intel Pentium processors all the same size

- 2.8 GHz tests out below 2.8 spec can be sold as a 2.66 GHz chip (down-
binning)
Principle of Nature
4-24
Trumpet of doom
As forecast horizon
increases, accuracy
decreases, e.g.,
Reduce production & delivery leadtimes

Dell pick-to-light system for assembly

Reduce information delays

EDI transmission of daily consumer demand up
through multiple levels in the supply chain
0
Forecast Error Range over Time
Time Until Forecast Event 0
Percentage
Forecast
Error
Principle of Nature
4-25
Reduce demand volatility
Beware of product proliferation

Pareto analysis separating the important few from the trivial many

Periodic length of line analysis to critically assess whether to continually
offer slow movers

Principle of Nature: Pareto phenomenon the lions share of an
aggregate measure is determined by relatively few factors

E.g., the 80-20 rule 80% of demand is due to 20% of product line

Beware of perverse cycle of promotions customers wait for
sale before buying, thereby forcing a sale

A step further dynamic pricing to stabilize demand & align with supply

Reduce the hockey stick effect
2 Principles of Nature
4-26
Hockey stick effect
Volume tends to pick up towards the end
of a reporting period . . . why?
Look for ways to lessen the effect
contributes to demand volatility,
inefficiency, poor service
Jan Feb
Principle of Nature
4-27
Channel stuffing
Lots of sales booked near the end of a quarter,
then sales drop off at the start of the next
quarter

E.g.,

A large brewer offered a vacation to the salesperson
in each region who sold the most beer to stores over
a 3 month period

One winner was able to convince a few stores to free
up backroom space and fill it entirely with beer
One contributor to the hockey stick effect
4-28
Were about to focus on
methods for predicting
demand
Improvement alternatives
short pork bellies
But, important to remember . . . many creative ways to
improve forecast accuracy that have nothing to do with
method

E.g., early warning incentives, law of large numbers, trumpet of
doom, reduce demand volatility
4-29
Road map
Processing Demand

Influencing Demand




Summary
4-30
Characteristics of long term forecasts
Single or multi-year horizon

Monthly or annual time bucket

Aggregate units

Input to long term decisions

Accuracy generally more important than short term
forecasts . . . why?

Tend to use expensive & time consuming methods . . .
due to the preceding point & due to a PON . . . which is?
4-31
Recency effect
Humans tend to overreact to (or be overly
influenced by) recent events

E.g.,

Hughes Electronics Corp. developed an artificial
intelligence based financial trading system. The
developers did this by encoding the wisdom of
Christine Downton, a successful portfolio manager.
One motivation for creating the system is that it is
immune to the recency effect, i.e., humans tend to
get overly fixated on the most recent information.

Principle of Nature
4-32
Some alternative methods
Judgment

Salesperson & customer input

Great information source, but beware of bias potential
& recency effect = humans tend to be overly
influenced by recent events

Outside services

Causal methods . . . examples?
4-33
Road map
Processing Demand

Influencing Demand




Characteristics

Components of demand

Moving average

Winters method

Focus forecasting

Filtering

Summary
4-34
Long term/short term characteristics
Long term forecasts

Single or multi-year horizon

Monthly or annual time bucket

Aggregate units (e.g., product/
service categories)

Input to long term decisions

Expensive & time consuming
methods
Accuracy importance
Trumpet of doom
Short term forecasts

Weekly or monthly horizon

Daily & weekly time bucket

Detailed units (e.g., SKU)

Input to short term decisions

Inexpensive & quick methods
Accuracy importance
Trumpet of doom
Could argue using 2 different principles of nature that its [easier?/harder?] to be
accurate with short term forecasting than with long term forecasting
4-35
Definition of the Forecasting Process
The Art and Science of Predicting Future
Events
Forecasting vs. Predicting
Based on Past Data
Economic vs. Demand Forecasting
4-36
Elements of Demand Forecasting
Dynamic in Nature
Consider Uncertainty (Stochastic)
Rely on Information contained in Past
Data
Applied to various time horizons
short term
medium term forecasts
long term forecasts
4-37
Steps in the Forecasting Process
Determine the Use of the Forecast
Select the Items to be Forecasted
Determine a Suitable Time Horizon
Select an appropriate Set of Forecasting Models
Gather Relevant Data
Conduct the Analysis
Validate the Model - Assess its Accuracy
Make the Forecast
Implement the Results
4-38
Independent Demand:
What a firm can do to manage it?
Can take an active role to influence
demand

FORECASTING
Can take a passive role and simply
respond to demand
4-39
Types of Forecasts
Qualitative (Judgmental)
Quantitative
Time Series Analysis
Causal Relationships
Simulation
4-40
Qualitative Methods
Grass Roots
Market Research
Panel Consensus
Executive Judgment
Historical analogy
Delphi Method
Qualitative
Methods
4-41
Delphi Method
1. Choose the experts to participate representing a
variety of knowledgeable people in different
areas
2. Through a questionnaire (or E-mail), obtain
forecasts (and any premises or qualifications for
the forecasts) from all participants
3. Summarize the results and redistribute them to
the participants along with appropriate new
questions
4. Summarize again, refining forecasts and
conditions, and again develop new questions
5. Repeat Step 4 as necessary and distribute the
final results to all participants
4-42
Quantitative Forecasting Models
Both Pattern Based and Correlational
Models rest on the assumption that the
relationships of the past will continue
into the Future
Both can Mathematically Characterize the
Probabilistic Nature of the Forecast
Both Use Information from Relevant
Time Frames
4-43
Road map
Processing Demand

Influencing Demand




Characteristics


Moving average

Winters method

Focus forecasting

Filtering

Summary
4-44
Components of Demand
Average demand for a period of
time
Trend
Seasonal element
Cyclical elements
Random variation
Autocorrelation
4-45
Pattern Based Analyses
Definition
Identifying an underlying pattern in historical
data, describe it in mathematical terms, and
then extrapolate it into the future
Uses a Time Series of Past Data
4-46
Time Series Variation
Time Series of Demand Data Typically
Contain Four Components of Variation
About the Mean or Average
Pattern Based Forecasting Needs to
Mathematically Characterize Each of
these
4-47
Finding Components of Demand
1 2 3 4
x
x
x
x
x
x
x x
x
x
x
x x x
x
x
x
x
x
x x
x
x
x
x x x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Year
S
a
l
e
s

Seasonal variation
Linear
Trend
Average
4-48
Time Series Analysis
Time series forecasting models try to
predict the future based on past data
You can pick models based on:
1. Time horizon to forecast
2. Data availability
3. Accuracy required
4. Size of forecasting budget
5. Availability of qualified personnel
4-49
Simple Moving Average Formula
F =
A + A + A +...+A
n
t
t -1 t -2 t -3 t - n
The simple moving average model assumes
an average is a good estimator of future
behavior
The formula for the simple moving average
is:
F
t
= Forecast for the coming period
n = Number of periods to be averaged
A
t-1
= Actual occurrence in the past period for up to n
periods

4-50
Simple Moving Average Problem (1)
Week Demand
1 650
2 678
3 720
4 785
5 859
6 920
7 850
8 758
9 892
10 920
11 789
12 844
F =
A + A + A +...+A
n
t
t -1 t -2 t -3 t - n
Question: What are the 3-
week and 6-week moving
average forecasts for
demand?
Assume you only have 3
weeks and 6 weeks of
actual demand data for the
respective forecasts
4-51
Week Demand 3-Week 6-Week
1 650
2 678
3 720
4 785 682.67
5 859 727.67
6 920 788.00
7 850 854.67 768.67
8 758 876.33 802.00
9 892 842.67 815.33
10 920 833.33 844.00
11 789 856.67 866.50
12 844 867.00 854.83
F
4
=(650+678+720)/3
=682.67
F
7
=(650+678+720
+785+859+920)/6
=768.67
Calculating the moving averages gives us:
4-52
500
600
700
800
900
1000
1 2 3 4 5 6 7 8 9 10 11 12
Week
D
e
m
a
n
d
Demand
3-Week
6-Week
Plotting the moving averages and comparing
them shows how the lines smooth out to reveal
the overall upward trend in this example
Note how the
3-Week is
smoother than
the Demand,
and 6-Week is
even smoother
4-53
Simple Moving Average Problem (2) Data
Week Demand
1 820
2 775
3 680
4 655
5 620
6 600
7 575
Question: What is the 3
week moving average
forecast for this data?
Assume you only have
3 weeks and 5 weeks
of actual demand
data for the
respective forecasts
4-54
Simple Moving Average Problem (2) Solution
Week Demand 3-Week 5-Week
1 820
2 775
3 680
4 655 758.33
5 620 703.33
6 600 651.67 710.00
7 575 625.00 666.00
F
4
=(820+775+680)/3
=758.33
F
6
=(820+775+680
+655+620)/5
=710.00
4-55
Weighted Moving Average Formula
F = w A + w A + w A +...+w A
t 1 t -1 2 t - 2 3 t -3 n t - n
w = 1
i
i=1
n
While the moving average formula implies an equal

weight being placed on each value that is being
averaged, the weighted moving average permits an
unequal weighting on prior time periods
w
t
= weight given to time period t
occurrence (weights must add to one)
The formula for the moving average is:
4-56
Weighted Moving Average Problem (1) Data
Weights:
t-1 .5
t-2 .3
t-3 .2
Week Demand
1 650
2 678
3 720
4
Question: Given the weekly demand and weights, what is
the forecast for the 4
th
period or Week 4?
Note that the weights place more emphasis on the
most recent data, that is time period t-1
4-57
Weighted Moving Average Problem (1) Solution
Week Demand Forecast
1 650
2 678
3 720
4 693.4
F
4
= 0.5(720)+0.3(678)+0.2(650)=693.4
4-58
Weighted Moving Average Problem (2) Data
Weights:
t-1 .7
t-2 .2
t-3 .1
Week Demand
1 820
2 775
3 680
4 655
Question: Given the weekly demand information and
weights, what is the weighted moving average forecast
of the 5
th
period or week?
4-59
Weighted Moving Average Problem (2) Solution
Week Demand Forecast
1 820
2 775
3 680
4 655
5 672
F
5
= (0.1)(755)+(0.2)(680)+(0.7)(655)= 672
4-60
Some pros/cons
1. Simple (+)

2. Designated weights of history (-)

3. History cut-off beyond m periods (-)
Short Term Forecasting Moving Average and Weighted Moving Average
4-61
Exponential Smoothing Model
Premise: The most recent observations might
have the highest predictive value
Therefore, we should give more weight to the
more recent time periods when forecasting
F
t
= F
t-1
+ o(A
t-1
- F
t-1
)
constant smoothing Alpha
period e past t tim in the occurance Actual A
period past time 1 in alue Forecast v F
period t time coming for the lue Forcast va F
: Where
1 - t
1 - t
t
=
=
=
=
o
4-62
Exponential Smoothing Problem (1) Data
Week Demand
1 820
2 775
3 680
4 655
5 750
6 802
7 798
8 689
9 775
10
Question: Given the
weekly demand data,
what are the
exponential smoothing
forecasts for periods 2-
10 using o=0.10 and
o=0.60?
Assume F
1
=D
1
4-63
Week Demand 0.1 0.6
1 820 820.00 820.00
2 775 820.00 820.00
3 680 815.50 793.00
4 655 801.95 725.20
5 750 787.26 683.08
6 802 783.53 723.23
7 798 785.38 770.49
8 689 786.64 787.00
9 775 776.88 728.20
10 776.69 756.28
Answer: The respective alphas columns denote the forecast
values. Note that you can only forecast one time period into
the future.
4-64
Exponential Smoothing Problem (1) Plotting
500
600
700
800
900
1 2 3 4 5 6 7 8 9 10
Week
D
e
m
a
n
d Demand
0.1
0.6
Note how that the smaller alpha results in a smoother line in
this example
4-65
Exponential Smoothing Problem (2) Data
Question: What are
the exponential
smoothing forecasts
for periods 2-5 using
a =0.5?
Assume F
1
=D
1
Week Demand
1 820
2 775
3 680
4 655
5
4-66
Exponential Smoothing Problem (2) Solution
Week Demand 0.5
1 820 820.00
2 775 820.00
3 680 797.50
4 655 738.75
5 696.88
F
1
=820+(0.5)(820-820)=820 F
3
=820+(0.5)(775-820)=797.75
4-67
Seasonal Adjustments
Applied to Moving Averages and Time
Series Regression
First, Calculate a Seasonal Index (SI)
Factor for Each Relevant Time Period
(day, week, month, quarter)
Each Seasonal Periods SI is
Calculated by Averaging the Ratio of
its Actual Demand to the Forecast
Demand for all Corresponding Periods
4-68
Forecast for Future Periods is Calculated
by Multiplying the Unadjusted Moving
Average or Time Series Forecast for a
given Period by the Corresponding
Seasonal Index for that Period
i.e. if the SMA forecast for the month of
March is 27 and the SI for March is
1.125, then
E
mar
= 27*1.125 = 30.375
4-69
Seasonal Adjustment Example
Sales Demand
Month 1993 1994
Monthly
Average
Overall
Average
Seasonal Index
SI Adjusted
Forecast
Jan 80 100 90.00 94.00 0.96 86.17
Feb 75 85 80.00 94.00 0.85 68.09
Mar 80 90 85.00 94.00 0.90 76.86
Apr 90 110 100.00 94.00 1.06 106.38
May 115 131 123.00 94.00 1.31 160.95
Jun 110 120 115.00 94.00 1.22 140.69
Jul 100 110 105.00 94.00 1.12 117.29
Aug 90 110 100.00 94.00 1.06 106.38
Sep 85 95 90.00 94.00 0.96 86.17
Oct 75 85 80.00 94.00 0.85 68.09
Nov 75 85 80.00 94.00 0.85 68.09
Dec 80 80 80.00 94.00 0.85 68.09
Average 87.92 100.08
Expected Demand for 1995 = 1153.23
4-70
Example Graph
Seasonal Adjusted Forecasting
50
70
90
110
130
150
170
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
1993
1994
SI Adjusted
Forecast
Overall
Average
4-71
Evaluating Forecast Accuracy
Use of Residuals Analyses
Residuals are the Difference Between the
Forecast and the Actual Demand for a Given
Period
Assessed by Several Measures
Mean Absolute Deviation - MAD
Mean Squared Error - MSE
Tracking Signal
4-72
The MAD Statistic to Determine
Forecasting Error
MAD =
A - F
n
t t
t=1
n
1 MAD 0.8 standard deviation

1 standard deviation 1.25 MAD
~
~
The ideal MAD is zero which would
mean there is no forecasting error

The larger the MAD, the less the
accurate the resulting model
4-73
MAD Problem Data
Month Sales Forecast
1 220 n/a
2 250 255
3 210 205
4 300 320
5 325 315
Question: What is the MAD value given
the forecast values in the table below?
4-74
MAD Problem Solution
MAD =
A - F
n
=
40
4
=10
t t
t=1
n
Month Sales Forecast Abs Error

1 220 n/a
2 250 255 5
3 210 205 5
4 300 320 20
5 325 315 10
40
Note that by itself, the MAD
only lets us know the mean
error in a set of forecasts
4-75
Mean Absolute Deviation - MAD
Exponentially Smoothed MAD
MAD
t
= o
MAD
|D
t
- Forecast
t
| + (1- o
MAD
)MAD
t-1
4-76
Mean Squared Error - MSE
MSE = (E(D
i
- Forecast
i
)
2
)/n
Period
Actual
Demand
Time
Series
Forecast
Time
Series
Residual
Squared
Error
1 12 12.16 -0.16 0.03
2 13 12.13 0.87 0.76
3 10 12.09 -2.09 4.39
4 11 12.06 -1.06 1.13
5 10 12.03 -2.03 4.12
6 14 12.00 2.00 4.01
7 16 11.97 4.03 16.28
8 15 11.93 3.07 9.40
9 13 11.90 1.10 1.21
10 8 11.87 -3.87 14.97
11 10 11.84 -1.84 3.37
12 12 11.80 0.20 0.04
13 9 11.77 -2.77 7.69
14 13 11.74 1.26 1.59
15 13 11.71 1.29 1.67
MSE = 4.71
RMSE = 2.17
4-77
Tracking Signal Formula
The Tracking Signal or TS is a measure that
indicates whether the forecast average is
keeping pace with any genuine upward or
downward changes in demand.
Depending on the number of MADs selected, the
TS can be used like a quality control chart
indicating when the model is generating too
much error in its forecasts.
The TS formula is:
TS =
RSFE
MAD
=
Running sum of forecast errors
Mean absolute deviation
4-78
Tracking Signal
Tracking Signal = Running Sum of
Forecast Error / MAD = RSFE/MAD
Period
Actual
Demand
Time
Series
Forecast
Time
Series
Residual
RSFE MAD
Tracking
Signal
1 12 12.16 -0.16 -0.16 0.03 -5.00
2 13 12.13 0.87 0.72 0.20 3.58
3 10 12.09 -2.09 -1.38 0.58 -2.38
4 11 12.06 -1.06 -2.44 0.68 -3.61
5 10 12.03 -2.03 -4.47 0.95 -4.72
6 14 12.00 2.00 -2.47 1.16 -2.13
7 16 11.97 4.03 1.57 1.73 0.90
8 15 11.93 3.07 4.63 2.00 2.32
9 13 11.90 1.10 5.73 1.82 3.15
10 8 11.87 -3.87 1.86 2.23 0.84
11 10 11.84 -1.84 0.03 2.15 0.01
12 12 11.80 0.20 0.22 1.76 0.13
13 9 11.77 -2.77 -2.55 1.96 -1.30
14 13 11.74 1.26 -1.29 1.82 -0.71
15 13 11.71 1.29 0.00 1.72 0.00
4-79
Road map
Processing Demand

Influencing Demand




Characteristics


Moving average

Winters method

Focus forecasting

Filtering

Summary
4-80
Old man winters
Winters method used to forecast one period into the future
See how method detects patterns & adapts to market changes over
time
Short Term Forecasting Winters
Old Man Winters in Action
0.00
100.00
200.00
300.00
400.00
500.00
600.00
0 20 40 60 80 100
Time
V
o
l
u
m
e
Actual
Forecast
4-81
Key to Winters method
Winters is an exponential smoothing
method

Smoothing is based on a key idea

For each component (which are?), a portion
of difference between estimate & actual is
due to randomness & certain portion due
to real change
4-82
Smoothing in action...
New estimate = old estimate + (some
percentage)(error)

Smoothes out peaks & valleys (i.e.,
randomness) of actual
4-83
Road map
Processing Demand

Influencing Demand




Characteristics


Moving average

Winters method

Focus forecasting

Filtering

Summary
4-84
Bernies insight
An intuitive & successful idea

Regularly use a # of different methods to
generate forecasts

Maintain historical accuracy information on each
method

Use the most accurate method to generate
official forecasts
Short Term Forecasting Focus
or what is focus forecasting?
4-85
Advertisement
appearing in
APICS The
Performance
Advantage
Short Term Forecasting Focus
4-86
Road map
Processing Demand

Influencing Demand




Characteristics


Moving average

Winters method

Focus forecasting

Filtering

Summary
4-87
Two types of filters
An important feature of computer-based forecasting
systems

Large amounts of data impractical to manually review all

1. For data input errors (e.g., typos, scanner errors)

If |actual - forecast| > limit, then report

2. For unacceptable forecast errors (e.g., warranting
management attention)

If average absolute error > limit, then report

If average error (i.e., bias) > limit, then report
Short Term Forecasting Filtering
4-88
Road map
Processing Demand
Influencing Demand
Dependent Demand
Correlational Forecasting
Summary
4-89
Demand Management
Bill of Materials (BOM)
A
B(4) C(2)
D(2) E(1) D(3) F(2)
Dependent Demand:
Raw Materials,
Component parts,
Sub-assemblies, etc.
Independent Demand:
Finished Goods
4-90
Web-Based Forecasting: CPFR
Collaborative Planning, Forecasting, and
Replenishment (CPFR) a Web-based tool used to
coordinate demand forecasting, production and
purchase planning, and inventory replenishment
between supply chain trading partners.
Used to integrate the multi-tier or n-Tier supply
chain, including manufacturers, distributors and
retailers.
CPFRs objective is to exchange selected internal
information to provide for a reliable, longer term
future views of demand in the supply chain.
CPFR uses a cyclic and iterative approach to
derive consensus forecasts.
4-91
Web-Based Forecasting:
Steps in CPFR
1. Creation of a front-end partnership
agreement
2. Joint business planning
3. Development of demand forecasts
4. Sharing forecasts
5. Inventory replenishment
4-92
Correlational Forecasting
Assumes an Outcome is Dependent an
Existing Relationship Between the
Demand Variable and Some other
Independent Variable(s)
Demand Variable is Dependent Variable
Other Related Variables are Independent
Variables
Generally Expressed as a Multiple Linear
Regression Model
Y = |
0
+ |
1
X
1
+ |
2
X
2
+ |
2
X
2
+ . . . |
n
X
n
+ c
i
4-93
Simple Linear Regression Model
Y
t
= a + bx
0 1 2 3 4 5 x (Time)
Y
The simple linear regression
model seeks to fit a line
through various data over
time
Is the linear regression model
a
- Y
t
is the regressed forecast value or dependent
variable in the model
-a is the intercept value of the the regression line, and
- b is similar to the slope of the regression line.
- However, since it is calculated with the variability of
the data in mind, its formulation is not as straight
forward as our usual notion of slope.
4-94
Simple Linear Regression Formulas for
Calculating a and b
a = y- bx
b =
xy- n(y)(x)
x - n(x
2 2
)
4-95
Simple Linear Regression Problem Data
Week Sales
1 150
2 157
3 162
4 166
5 177
Question: Given the data below, what is the simple linear
regression model that can be used to predict sales in future
weeks?
4-96
Week Week*Week Sales Week*Sales
1 1 150 150
2 4 157 314
3 9 162 486
4 16 166 664
5 25 177 885
3 55 162.4 2499
Average Sum Average Sum
b =
xy- n( y)(x)
x - n(x
=
2499- 5(162.4)(3)
=
a = y- bx =162.4 - (6.3)(3) =
2 2

=
) ( ) 55 5 9
63
10
6.3
143.5
Answer: First, using the linear regression formulas, we
can compute a and b
4-97
Y
t
= 143.5 + 6.3x
180
Perio
d
135
140
145
150
155
160
165
170
175
1 2 3 4 5
S
a
l
e
s

Sales
Forecast
The resulting regression model
is:
Now if we plot the regression generated forecasts against the
actual sales we obtain the following chart:
97
4-98
Statistical Assumptions of Multiple Linear
Regression
The Error Term (the residual c
i
) is
Normally Distributed
There is no Serial Correlation Among
Error Terms
Magnitude of the Error Term is
Independent of the Size of Any of the
Independent Variables - X
i

Assumptions Can be Tested Through
Analyses of the Residuals - c
i
4-99
Major Statistical Problems of Multiple
Linear Regression
Multicolinarity
Use of Time-Lagged Independent
Variables
Both of These Problems Result in Models
with Potentially Valid Predictions, but the
Reliability of the | Coefficients is
Questionable
McGraw-Hill/Irwin
Copyright 2008 by The McGraw-Hill Companies, Inc. All rights reserved.
Demand Management
The End
Processing,
Influencing, &
Anticipating
Demand
Buy
Sell
Buy
Store
Move
Make
Sell
Store
Move
Sell
Make
Make
Move
Move
Buy

Demand Forecasting

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Demand Forecasting

Uploaded by

Copyright:

Available Formats

McGraw-Hill/Irwin

Copyright 2008 by The McGraw-Hill Companies, Inc. All rights reserved.

While the moving average formula implies an equal

1 MAD 0.8 standard deviation

Month Sales Forecast Abs Error

You might also like