Operations Management

Chapter 3
Forecasting

Forecasting Provides a Competitive

Advantage for Disney (1 of 4)

• Global portfolio includes parks in Shanghai, Hong Kong,

Paris, Tokyo, Orlando, and Anaheim
• Revenues are derived from people – how many visitors
and how they spend their money
• Daily management report contains only the forecast and
actual attendance at each park

2
 Forecasting Provides a Competitive
Advantage for Disney (2 of 4)

• Disney generates daily, weekly, monthly, annual, and 5-

year forecasts
• Forecast used by labor management, maintenance,
operations, finance, and park scheduling
• Forecast used to adjust opening times, rides, shows,
staffing levels, and guests admitted

Forecasting Provides a Competitive

Advantage for Disney (3 of 4)

• 20% of customers come from outside the USA

• Economic model includes gross domestic product, cross-
exchange rates, arrivals into the USA
• A staff of 35 analysts and 70 field people survey 1 million
park guests, employees, and travel professionals each
year

4
 Forecasting Provides a Competitive
Advantage for Disney (4 of 4)

• Inputs to the forecasting model include airline specials,

Federal Reserve policies, Wall Street trends,
vacation/holiday schedules for 3,000 school districts
around the world
• Average forecast error for the 5-year forecast is 5%
• Average forecast error for annual forecasts is between
0% and 3%

Learning Objectives (1 of 2)

4.1 Understand the three time horizons and which

models apply for each
4.2 Explain when to use each of the four qualitative
models
4.3 Apply the naive, moving-average, exponential
smoothing, and trend methods

6
 Learning Objectives (2 of 2)

4.4 Compute three measures of forecast accuracy

4.5 Develop seasonal indices
4.6 Conduct a regression and correlation analysis
4.7 Use a tracking signal

What Is Forecasting?
• Process of predicting a
future event
• Underlying basis of all
business decisions
– Production
– Inventory
– Personnel
– Facilities

8
 Forecasting Time Horizons
1. Short-range forecast
– Up to 1 year, generally less than 3 months
– Purchasing, job scheduling, workforce levels, job
assignments, production levels
2. Medium-range forecast
– 3 months to 3 years
– Sales and production planning, budgeting
3. Long-range forecast
– 3+ years
– New product planning, facility location, research and
development

Distinguishing Differences

1. Medium/long range forecasts deal with more

comprehensive issues and support management
decisions regarding planning and products, plants and
processes
2. Short-term forecasting usually employs different
methodologies than longer-term forecasting
3. Short-term forecasts tend to be more accurate than
longer-term forecasts

10
 Influence of Product Life Cycle

Introduction – Growth – Maturity – Decline

• Introduction and growth require longer forecasts than
maturity and decline
• As product passes through life cycle, forecasts are useful
in projecting
– Staffing levels
– Inventory levels
– Factory capacity

11

Figure 2.5 Product Life Cycle (1 of 2)

12
 Figure 2.5 Product Life Cycle (2 of 2)

13

Types of Forecasts

1. Economic forecasts
– Address business cycle – inflation rate, money
supply, housing starts, etc.
2. Technological forecasts
– Predict rate of technological progress
– Impacts development of new products
3. Demand forecasts

– Predict sales of existing products and services

14
 Strategic Importance of Forecasting

• Supply-Chain Management – Good supplier relations,

advantages in product innovation, cost and speed to
market
• Human Resources – Hiring, training, laying off workers
• Capacity – Capacity shortages can result in
undependable delivery, loss of customers, loss of market
share

15

Seven Steps in Forecasting

1. Determine the use of the forecast

2. Select the items to be forecasted
3. Determine the time horizon of the forecast
4. Select the forecasting model(s)
5. Gather the data needed to make the forecast
6. Make the forecast
7. Validate and implement the results

16
 The Realities!

• Forecasts are seldom perfect, unpredictable outside

factors may impact the forecast
• Most techniques assume an underlying stability in the
system
• Product family and aggregated forecasts are more
accurate than individual product forecasts

17

Forecasting Approaches (1 of 2)

Qualitative Methods
• Used when situation is vague and little data exist
– New products
– New technology
• Involves intuition, experience
– e.g., forecasting sales on Internet

18
 Forecasting Approaches (2 of 2)

Quantitative Methods
• Used when situation is ‘stable’ and historical data exist
– Existing products
– Current technology
• Involves mathematical techniques
– e.g., forecasting sales of color televisions

19

Overview of Qualitative Methods (1 of 2)

1. Jury of executive opinion

– Pool opinions of high-level experts, sometimes

augmented by statistical models
2. Delphi method
– Panel of experts, queried iteratively

20
 Overview of Qualitative Methods (2 of 2)

3. Sales force composite

– Estimates from individual salespersons are reviewed
for reasonableness, then aggregated
4. Market Survey
– Ask the customer

21

Jury of Executive Opinion

• Involves small group of high-level experts and managers

• Group estimates demand by working together
• Combines managerial experience with statistical models
• Relatively quick
• ‘Group-think’ disadvantage

22
 Delphi Method

• Iterative group process,

continues until consensus is
reached
• Three types of participants
– Decision makers
– Staff
– Respondents

23

Sales Force Composite

• Each salesperson projects his or her sales

• Combined at district and national levels
• Sales reps know customers’ wants
• May be overly optimistic

24
 Market Survey

• Ask customers about purchasing plans

• Useful for demand and product design and planning
• What consumers say and what they actually do may be
different
• May be overly optimistic

25

Overview of Quantitative Approaches

1. Naive approach
2. Moving averages
3. Exponential smoothing
4. Trend projection
5. Linear regression

26
 Time-Series Forecasting

• Set of evenly spaced numerical data

– Obtained by observing response variable at regular
time periods
• Forecast based only on past values, no other variables
important
– Assumes that factors influencing past and present will
continue influence in future

27

Time-Series Components

28
 Components of Demand
Figure 4.1 Demand Charted over 4 Years, with a Growth Trend and
Seasonality Indicated

29

Trend Component

• Persistent, overall upward or downward pattern

• Changes due to population, technology, age, culture, etc.
• Typically several years duration

30
 Seasonal Component

• Regular pattern of up and down fluctuations

• Due to weather, customs, etc.
• Occurs within a single year

Period Length “Season” Length Number Of “Seasons” In Pattern

Week Day 7
Month Week 4 – 4.5
Month Day 28 – 31
Year Quarter 4
Year Month 12
Year Week 52

31

Cyclical Component

• Repeating up and down

movements
• Affected by business
cycle, political, and
economic factors
• Multiple years duration
• Often causal or
associative relationships

32
 Random Component

• Erratic, unsystematic, ‘residual’ fluctuations

• Due to random variation or unforeseen events
• Short duration and nonrepeating

33

Naive Approach

• Assumes demand in next period is the

same as demand in most recent period
– e.g., If January sales were 68, then
February sales will be 68
• Sometimes cost effective and efficient
• Can be good starting point

34
 Moving Averages

• MA is a series of arithmetic means

• Used if little or no trend
• Used often for smoothing
– Provides overall impression of data over time

Moving average =
 demand in previous n periods
n

35

Moving Average Example

36
 Weighted Moving Average (1 of 3)

• Used when some trend might be present

– Older data usually less important
• Weights based on experience and intuition

moving = 
Weighted
( ( Weight for period n )(Demand in period n ) )
average  Weights

37

Weighted Moving Average (2 of 3)

38
 Weighted Moving Average (3 of 3)

39

Potential Problems with Moving Average

1. Increasing n smooths the forecast but makes it less

sensitive to changes
2. Does not forecast trends well
3. Requires extensive historical data

40
 Graph of Moving Averages
Figure 4.2 Actual Demand vs. Moving-Average and Weighted-Moving-
Average Methods for Donna’s Garden Supply

41

Exponential Smoothing (1 of 2)

• Form of weighted moving average

– Weights decline exponentially
– Most recent data weighted most
• Requires smoothing constant (α)
– Ranges from 0 to 1
– Subjectively chosen
• Involves little record keeping of past data

42
 Exponential Smoothing (2 of 2)

New forecast = Last period’s forecast + α (Last period’s

actual demand − Last period’s forecast)

Ft = Ft −1 +  ( A t −1 − Ft −1 )

where Ft = new forecast

Ft – 1 = previous period’s forecast
α = smoothing (or weighting) constant ( 0    1)
At – 1 = previous period’s actual demand

43

Exponential Smoothing Example (1 of 3)

Predicted demand = 142 Ford Mustangs

Actual demand = 153
Smoothing constant α = .20

44
 Exponential Smoothing Example (2 of 3)

Predicted demand = 142 Ford Mustangs

Actual demand = 153
Smoothing constant α = .20

New forecast = 142 + .2(153 − 142)

45

Exponential Smoothing Example (3 of 3)

Predicted demand = 142 Ford Mustangs

Actual demand = 153
Smoothing constant α = .20

New forecast = 142 + .2(153 − 142)

= 142 + 2.2
= 144.2  144 cars

46
 Effect of Smoothing Constants

• Smoothing constant generally .05  α  .50

• As α increases, older values become less significant

47

Impact of Different  (1 of 2)

48
 Impact of Different  (2 of 2)

• Choose high values of  when underlying average is

likely to change
• Choose low values of  when underlying average is
stable

49

Selecting the Smoothing Constant

The objective is to obtain the most accurate forecast no

matter the technique
We generally do this by selecting the model that gives
us the lowest forecast error according to one of three
preferred measures:
• Mean Absolute Deviation (MAD)
• Mean Squared Error (MSE)
• Mean Absolute Percent Error (MAPE)

50
 Common Measures of Error

Mean Absolute Deviation (MAD)

MAD =
 | Actual − Forecast
n

51

Determining the MAD (1 of 2)

Actual
Tonnage Forecast With
Quarter Unloaded Forecast With a = .10 a = .50
1 180 175 175
2 168 175.50 = 175.00 + .10(180 − 175) 177.50
3 159 174.75 = 175.50 + .10(168 − 175.50) 172.75
4 175 173.18 = 174.75 + .10(159 − 174.75) 165.88
5 190 173.36 = 173.18 + .10(175 − 173.18) 170.44
6 205 175.02 = 173.36 + .10(190 − 173.36) 180.22
7 180 178.02 = 175.02 + .10(205 − 175.02) 192.61
8 182 178.22 = 178.02 + .10(180 − 178.02) 186.30
9 ? 178.59 = 178.22 + .10(182 − 178.22) 184.15

52
 Determining the MAD (2 of 2)

53

Common Measures of Error (1 of 2)

Mean Squared Error (MS)

 (Forecast errors )
2

MSE =
n

54
 Determining the MSE

 (Forecast errors )
2

MSE= = 1,526.52 8 = 190.8

n

55

Common Measures of Error (2 of 2)

Mean Absolute Percent Error (MAPE)

100 Actual
i =1
i − Forecast i Actuali
MAPE=
n

56
 Determining the MAPE

MAPE=
 absolute percent error = 44.75% = 5.59%
n 8

57

Comparison of Measures

Table 4.1 Comparison of Measures of Forecast Error

Measure Meaning Application To Chapter Example

Mean absolute How much the For a = .10 in Example 4, the forecast for grain
deviation (MA forecast missed the unloaded was off by an average of 10.31 tons.
D) target
Mean squared The square of how For a = .10 in Example 5, the square of the
error (MSE) much the forecast forecast error was 190.8. This number does not
missed the target have a physical meaning, but is useful when
compared to the MSE of another forecast.
Mean absolute The average percent For a = .10 in Example 6, the forecast is off by
percent error error 5.59% on average. As in Examples 4 and 5,
(MAPE) some forecasts were too high, and some were
low.

58
 Comparison of Forecast Error (1 of 7)
Actual Rounded Absolute Rounded Absolute
Tonnage Forecast Deviation for Forecast Deviation for
Quarter Unloaded with α = .10 α = .10 with α = .50 α = .50
1 180 175 5.00 175 5.00
2 168 175.5 7.50 177.50 9.50
3 159 174.75 15.75 172.75 13.75
4 175 173.18 1.82 165.88 9.12
5 190 173.36 16.64 170.44 19.56
6 205 175.02 29.98 180.22 24.78
7 180 178.02 1.98 192.16 12.61
8 182 178.22 3.78 186.30 4.30
Blank Blank Blank 82.45 Blank 98.62

59

Comparison of Forecast Error (2 of 7)

MAD =
 deviations
n

For  = .10
= 82.45 8 = 10.31

For  = .50
= 98.62 8 = 12.33

60
 Comparison of Forecast Error (3 of 7)

61

Comparison of Forecast Error (4 of 7)

 ( forecast errors )
2

MSE =
n

For α = .10
= 1,526.52 8 = 190.8
For α = .50
= 1,561.91 8 = 195.24

62
 Comparison of Forecast Error (5 of 7)

63

Comparison of Forecast Error (6 of 7)

n

100 deviation i actuali

MAPE = i =1

For α = .10
= 44,75% 8 = 5.59%
For α = .50
= 54,00% 8 = 6.75%

64
 Comparison of Forecast Error (7 of 7)

65

Exponential Smoothing with Trend

Adjustment (1 of 3)

When a trend is present, exponential smoothing must be

modified

Month Actual Demand Forecast (Ft) For Months 1 − 5

1 100 F1 = 100 (given)

2 200 F2 = F1 + α(A1 − F1) = 100 + .4(100 − 100) = 100

3 300 F3 = F2 + α (A2 − F2) = 100 + .4(200 − 100) = 140

4 400 F4 = F3 + α (A3 − F3) = 140 + .4(300 − 140) = 204

5 500 F5 = F4 + α (A4 − F4) = 204 + .4(400 − 204) = 282

66
 Exponential Smoothing with Trend
Adjustment (2 of 3)
Forecast including trend (FITt) = Exponentially smoothed
forecast (Ft) + Exponentially smoothed trend (Tt)
Ft = α ( At −1 ) + (1− α )( Ft −1 + Tt −1 )

Tt = β ( Ft − Ft −1 ) + (1− β )Tt −1

Where Ft = exponentially smoothed forecast average

Tt = exponentially smoothed trend
At = actual demand
α = smoothing constant for average ( 0  α  1)
β = smoothing constant for trend ( 0  β  1)

67

Exponential Smoothing with Trend

Adjustment (3 of 3)

Step 1: Compute Ft
Step 2: Compute Tt
Step 3: Calculate the forecast FITt = Ft + Tt

68
 Exponential Smoothing with Trend
Adjustment Example (1 of 6)

Month (t) Actual Demand (At) Month (t) Actual Demand (At)
1 12 6 21
2 17 7 31
3 20 8 28
4 19 9 36
5 24 10 ?

α = .2 β = .4

69

Exponential Smoothing with Trend

Adjustment Example (2 of 6)

Table 4.2 Forecast with α = .2 and β = .4

70
 Exponential Smoothing with Trend
Adjustment Example (3 of 6)

71

Exponential Smoothing with Trend

Adjustment Example (4 of 6)

72
 Exponential Smoothing with Trend
Adjustment Example (5 of 6)

Smoothed
Forecast Average, Smoothed Forecast Including
Month Actual Demand Ft Trend, Tt Trend, Fitt
1 12 11 2 13.00
2 17 12.80 1.92 14.72
3 20 15.18 2.10 17.28
4 19 17.82 2.32 20.14
5 24 19.91 2.23 22.14
6 21 22.51 2.38 24.89
7 31 24.11 2.07 26.18
8 28 27.14 2.45 29.59
9 36 29.28 2.32 31.60
10 — 32.48 2.68 35.16

73

Exponential Smoothing with Trend

Adjustment Example (6 of 6)
Figure 4.3 Exponential Smoothing with Trend-Adjustment Forecasts
Compared to Actual Demand Data

74
 Trend Projections

Fitting a trend line to historical data points to project into

the medium to long-range
Linear trends can be found using the least squares
technique
Λ
y = a + bx
Λ
where y = computed value of the variable to
be predicted (dependent variable)
a = y-axis intercept
b = slope of the regression line
x = the independent variable

75

Least Squares Method (1 of 2)

Figure 4.4 The Least-Squares Method for Finding the Best-Fitting
Straight Line, Where the Asterisks Are the Locations of the Seven
Actual Observations or Data Points

Least squares method minimizes the sum of the squared errors

(deviations)

76
 Least Squares Method (2 of 2)

• Equations to calculate the regression variables

ŷ = a + bx

b=
å xy - nxy
å x - nx
2 2

a = y - bx

77

Least Squares Example (1 of 4)

Electrical Electrical
Year Power Demand Year Power Demand
1 74 5 105
2 79 6 142
3 80 7 122
4 90 Blank Blank

78
 Least Squares Example (2 of 4)

Electrical Power
Year (x) x2 xy
Demand (y)
1 74 1 74
2 79 4 158
3 80 9 240
4 90 16 360
5 105 25 525
6 142 36 852
7 122 49 854
 = x28
x of
sum = 28 692.
sum of y = y = 692 sum = 140=
x 2squared
ofx  ofxyxy == 3,063
sum 3,063
140.

x=
 x = 28 = 4 y=
 y = 692 = 98.86
n 7 n 7

79

Least Squares Example (3 of 4)

b=
 xy − nxy = 3,063 − ( 7 )( 4 )( 98.86 ) = 295 = 10.54
 x − nx 140 − ( 7 ) ( 4 )
2 2
2 28

a = y − bx = 98.86 − 10.54 ( 4 ) = 56.70


Thus, y = 56.70 + 10.54 x

Demand in year 8 = 56.70 + 10.54(8)

= 141.02, or 141 megawatts

80
 Least Squares Example (4 of 4)
Figure 4.5 Electrical Power and the Computed Trend Line

81

Least Squares Requirements

1. We always plot the data to insure a linear relationship

2. We do not predict time periods far beyond the database
3. Deviations around the least squares line are assumed to
be random

82
 Seasonal Variations in Data (1 of 2)

The multiplicative seasonal model can adjust trend data for

seasonal variations in demand

83

Seasonal Variations in Data (2 of 2)

Steps in the process for monthly seasons:

1. Find average historical demand for each month
2. Compute the average demand over all months
3. Compute a seasonal index for each month
4. Estimate next year’s total demand
5. Divide this estimate of total demand by the number of
months, then multiply it by the seasonal index for that
month

84
 Seasonal Index Example (1 of 6)

85

Seasonal Index Example (2 of 6)

1,128
Average monthly demand = = 94
12 months
Copyright © 2017, 2014, 2011 Pearson Education, Inc. All Rights Reserved

86
 Seasonal Index Example (3 of 6)

Average monthly demand for past 3 years

Seasonal index =
Average monthly demand

87

Seasonal Index Example (4 of 6)

88
 Seasonal Index Example (5 of 6)

Seasonal forecast for Year 4

Month Demand Month Demand
1,200 over 12, end fraction, times 0.957 = 96 1,200 over 12, end fraction, times 1.117 = 112

Jan 1,200
 .957 = 96 July 1,200
 1.117 = 112
12 12
1,200 over 12, end fraction, times 0.851 = 85 1,200 over 12, end fraction, times 1.064 = 106

Feb 1,200 Aug 1,200

 .851 = 85  1.064 = 106
12 12
1,200 over 12, end fraction, times 0.904 = 90 1,200 over 12, end fraction, times 0.957 = 96

Mar 1,200
 .904 = 90
Sept 1,200
 .957 = 96
12 12
1,200 over 12, end fraction, times 1.064 = 106 1,200 over 12, end fraction, times 0.851 = 85

Apr 1,200 Oct 1,200

 1.064 = 106  .851 = 85
12 12
1,200 over 12, end fraction, times 1.309 = 131 1,200 over 12, end fraction, times0.851 = 85

May 1,200 Nov 1,200

 1.309 = 131  .851 = 85
12 12
1,200 over 12, end fraction, times 1.223 = 122 1,200 over 12, end fraction, times0.851 = 85

June 1,200 Dec 1,200

 1.223 = 122  .851 = 85
12 12

89

Seasonal Index Example (6 of 6)

90
 San Diego Hospital (1 of 5)
Figure 4.6 Trend Data for San Diego Hospital

91

San Diego Hospital (2 of 5)

Seasonality Indices for Adult Inpatient Days at San Diego

Hospital

Month Seasonality Index Month Seasonality Index

January 1.04 July 1.03
February 0.97 August 1.04
March 1.02 September 0.97
April 1.01 October 1.00
May 0.99 November 0.96
June 0.99 December 0.98

92
 San Diego Hospital (3 of 5)
Figure 4.7 Seasonal Index for San Diego Hospital

93

San Diego Hospital (4 of 5)

Period 67 68 69 70 71 72
Month Jan Feb Mar Apr May June
Forecast with 9,911 9,265 9,764 9,691 9,520 9,542
Trend &
Seasonality
Period 73 74 75 76 77 78
Month July Aug Sept Oct Nov Dec
Forecast with 9,949 10,068 9,411 9,724 9,355 9,572
Trend &
Seasonality

94
 San Diego Hospital (5 of 5)
Figure 4.8 Combined Trend and Seasonal Forecast

95

Adjusting Trend Data

yˆ seasonal = Index  yˆ trendforcast

Quarter I: yˆI = (1.30)($100,000) = $130,000

Quarter II: yˆII = (.90)($120,000) = $108,000

Quarter III: yˆIII = (.70)($140,000) = $98,000

Quarter IV: yˆIV = (1.10)($160,000) = $176,000

96
 Cyclical Variations

• Cycles – patterns in the data that occur every several

years
– Forecasting is difficult
– Wide variety of factors

97

Associative Forecasting (1 of 2)

Used when changes in one or more independent variables

can be used to predict the changes in the dependent
variable
Most common technique is linear-regression analysis
We apply this technique just as we did in the time-
series example

98
 Associative Forecasting (2 of 2)

Forecasting an outcome based on predictor variables using

the least squares technique
Λ
y = a + bx

Λ
where y = value of the dependent variable (in our example,
sales)
a = y-axis intercept
b = slope of the regression line
x = the independent variable

99

Associative Forecasting Example (1 of 6)

Nodel’s Sales Area Payroll Nodel’s Sales Area Payroll
(In $ Millions), y (In $ Billions), x (In $ Millions), y (In $ Billions), x
2.0 1 2.0 2
3.0 3 2.0 1
2.5 4 3.5 7

100
 Associative Forecasting Example (2 of 6)
Sales, y Payroll, x x2 xy
2.0 1 1 2.0
3.0 3 9 9.0
2.5 4 16 10.0
2.0 2 4 4.0
2.0 1 1 2.0
3.5 7 49 24.5
the sum of Y= 15.0 the sum of X = 18 the sum of X squared = 18 the sum of X Y = 51.5

 y = 15.0  x = 18  x 2 = 18  xy = 51.5

x=
 x = 18 = 3 y=
 y = 15 = 2.5
6 6 6 6

b=
 xy − nxy = 51.5 − ( 6 )( 3 )( 2.5 ) = .25 a = y − bx = 2.5 − (.25 )( 3 ) = 1.75
 x − nx 80 − ( 6 ) ( 3 )
2 2 2

101

Associative Forecasting Example (3 of 6)


y = 1.75 + .25 x
Sales = 1.75 + .25 (payroll )

102
 Associative Forecasting Example (4 of 6)

103

Associative Forecasting Example (5 of 6)

If payroll next year is estimated to be $6 billion, then:

Sales ( in $ millions ) = 1.75 + .25 ( 6 )

= 1.75 + 1.5 = 3.25
Sales = $3,250,000

104
 Associative Forecasting Example (6 of 6)

105

Fast Food Restaurant Forecast

Figure 4.12 Forecasts Are Unique: Note the Variations between (a)
Hourly Sales at a Fast-Food Restaurant and (b) Hourly Call Volume at
FedEx

124
 Fedex Call Center Forecast
Figure 4.12 [continued]

125