FORECASTING

z
DEMAND
Module 4
What is Forecasting?
Forecasting is the process of making predictions of the
future based on past and present data.
❖ A forecast is an estimate of what is likely to happen in the
future.
z
✓ The primary advantage of forecasting is that it provides the business
with valuable information that the business can use to make decisions
about the future of the organization. (Bass, 2018)
Decisions Requiring Forecasting in Operations
Management

❖ Predicting demands of new and existing products

❖ Predicting
z results of new product research and
development
❖ Projecting quality improvement
❖ Anticipating customer’s needs
❖ Predicting cost of materials
What is Demand Forecast?
Demand forecasting is the process of making estimations
about future customer demand over a defined period, using
historical data and other information.
z
Decisions Relevant to Demand Forecasts
❖ Select product portfolio
❖ Predicting new facility location
❖ Anticipating capacity needs
z ❖ Identifying labor requirements
❖ Projecting material requirements
❖ Developing production schedules
❖ Creating maintenance schedules
 Forecasting Time Horizons
1. Short range forecast
Has time span of up to 1 year but less than 3
months
z used for:
➢ planning purchases
➢ Hiring
➢ job assignments
➢ production level
 Forecasting Time Horizons
2. Medium range forecast
Three months to three years
Used for:
z➢ Sales and production planning
➢ Budgeting
➢ Analysis of different operating plans
 Forecasting Time Horizons
3. Long range forecast
Three years or more
Used for:
z➢ New products
➢ Capital expenditures
➢ Facility expansion
➢ Relocation
➢ Research and development
 Types of Forecast
1. Economic forecasts: Make
predictions related to inflation,
money supplies, and other
economic factors that can
z
affect businesses. These
forecasts often influence
medium to long range
planning
 Types of Forecast
2. Technological forecasts:
monitor rates of
technological progress. This
keepszorganizations abreast
of trends and can result in
exciting new products. New
products may require new
facilities and equipment,
which must be planned for in
the appropriate time frame.
 Types of Forecast
3. Demand forecasts:
Estimate consumer demand
for a business' products or
services.
z A demand forecast
will be used to estimate
production and all relevant
inputs. These forecasts can
inform short, medium, and
long term planning. Also
referred to as sales forecasts.
 The Strategic Importance of Forecasting
1. Supply Chain Management
- rapid communication and accurate data shared up and down to
supply chain innovation is enhanced
- inventoryz
cost are reduced
- improved respond in market demand and supply
- collaborate with suppliers to make sure the right items is available
at the right time in the right place and the right price.
- efficient supply chain and satisfied customers

❖ CPFR – Collaborative planning, forecasting, and replenishment

✓ The goal of CPFR is to create significantly more accurate information that
can power the supply chain to greater sales and profits
 The Strategic Importance of Forecasting
2. Human Resources
- Workforce is based on demand
- Hiring, training and laying off depend on anticipated demand.
z
3. Capacity
- When the capacity cannot keep up to the demand, the result
is undependable delivery, loss of customers, and maybe loss
of market share.
 Steps in Forecasting Process
Determine
Determine the time Gather data
the purpose horizon of needed to Validate and
of the the make the implement
forecast forecast forecast the results

1 z2 3 4 5 6 7

Select the Select the Make the

items to be forecasting forecast
forecasted model(s)
 FORECASTING
z
APPROACHES
 Two Approaches in Forecasting
1. Qualitative Forecasts Method
- A qualitative approach uses
factors such as experience,
instinct and emotion inventory
cost are reduced
z
2. Quantitative Forecasts Method
- Forecasts that employ
mathematical modeling to
forecast demand.
Four Different Qualitative Forecasting Techniques
Qualitative Forecasts Methods
Qualitative Methods

z
Executive Opinion
--------------------------
Approach in which
a group of
managers meet
and collectively
develop a forecast
Sales Force
Market Survey Composite Delphi Method
-------------------------- -------------------------- --------------------------
Approach that uses
interviews and Approach in which Approach in which
surveys to judge each salesperson consensus
preferences of estimates sales in agreement is
customers and to their region reached among a
assess demand group of experts
Two Categories of Quantitative Forecast Methods
Quantitative Forecast Methods
Quantitative Methods
Time-Series Models
-------------------------------------------------
Time series models look at past
patterns of data and attempt to
predict the future based upon the
underlying patterns contained within
those data.
Associative Models
-------------------------------------------------
Associative models assume that the
variable being forecasted is related
to other variable in the environment.
They try to project based upon those
associations.
 Quantitative
z
Forecast
Time Series Models
 Quantitative Forecast Method
Time Series Models – predict on the assumption that the future
is a function of the past.

Four components
z of Time Series Models:
1. Trend
2. Seasonality
3. Cycles
4. Random Variations
 Quantitative Forecast Method
Trend – data exhibits an
increasing or decreasing
pattern
z
❖ A trend is consistent upward
or downward movement of the
demand. This may be related to
the product’s life cycle.
 Quantitative Forecast Method
Seasonality – any pattern that
regularly repeats itself and is
of a constant length.
❖ Many products
z have a seasonal
pattern, generally predictable
changes in demand that are
recurring every year. Fashion
products and sporting goods are
heavily influenced by
seasonality.
 Quantitative Forecast Method
Cycle – patterns created by
economic fluctuations

❖ A cycle isza pattern in the data

that tends to last more than one
year in duration. Often, they are
related to events such as interest
rates, the political climate,
consumer confidence or other
market factors.
 Quantitative Forecast Method
Random Variations – Random variations are the unexplained
variations in demand that remain after all other factors are
considered. Often this is referred to as noise.
z
Irregular Variations - Often demand can be influenced by an
event or series of events that are not expected to be repeated in
the future. Examples might include an extreme weather event, a
strike at a college campus, or a power outage.
 Qualitative Forecast Method (Time Series Methods)
1. Naïve Method– The simplest forecasting method is the naïve
method. In this case, the forecast for the next period is set at
the actual demand for the previous period. This method of
forecastingzmay often be used as a benchmark in order to
evaluate and compare other forecast methods.
Qualitative Forecast Method (Time Series Methods)
1. Naïve Method (Cookies)
Day Actual Forecast
Demand (At) (Ft)
1 17
2 18 17
z 3 22 18
4 20 22
5 19 20
6 16 19
7 15 16
8 20 15
9 21 20
10 18 21
11 17 18
 Qualitative Forecast Method (Time Series Methods)
2. Simple Moving Average
- The average value over a set time period (e.g. the last three
weeks or months or year)
❖ Mathematically,
z the simple moving average (which serves
as an estimate of the next period’s demand) is expressed
as:
∑Actual demand in previous n periods
Moving Average = ---------------------------------------------------------
n
where n is the number of periods in the moving average
 Qualitative Forecast Method (Time Series Methods)
Months ACTUAL 3-month
DEMAND moving
2. Simple Moving Average (At) average
❖ Forecast for three months 1 310 -
2 360 -
3 380 -
z 4 415 350
Moving average = 310 + 360 + 380 5 450 385
3 6 470 415
Moving average = 1050 7 445
3
Moving average= 350
 Qualitative Forecast Method (Time Series Methods)
Months ACTUAL 3-month
DEMAND moving
2. Simple Moving Average (At) average
❖ Forecast for three months 1 310 -
2 360 -
3 380 -
z 4 415 350
Moving average = 310 + 360 + 380 5 450 385
3 6 470 415
Moving average = 1050 7 445
3
Moving average= 350
 Qualitative Forecast Method (Time Series Methods)
Months ACTUAL 3-month
DEMAND moving
2. Simple Moving Average (At) average
❖ Forecast for three months 1 310 -
2 360 -
3 380 -
z 4 415 350
Moving average = 360 + 380 + 415 5 450 385
3 6 470 415
Moving average = 1155 7 445
3
Moving average= 385
 Qualitative Forecast Method (Time Series Methods)
Month ACTUAL 3-month
DEMAND moving
2. Simple Moving Average (At) average
❖ Forecast for three months 1 310 -
2 360 -
3 380 -
z 4 415 350
Moving average = 380 + 415 + 450 5 450 385
3 6 470 415
Moving average = 1245 7 445
3
Moving average= 415
 Qualitative Forecast Method (Time Series Methods)
Month ACTUAL 3-month
DEMAND moving
2. Simple Moving Average (At) average
❖ Forecast for three months 1 310 -
2 360 -
3 380 -
z 4 415 350
Moving average = 415 + 450 + 470 5 450 385
3 6 470 415
Moving average = 1335 7 445
3
Moving average= 445
 Qualitative Forecast Method (Time Series Methods)
3. Weighted Moving Average
- This method is the same as the simple moving average with
the addition of a weight for each one of the last “n” periods. In
practice,
z these weights need to be determined in a way to
produce the most accurate forecast.
- more responsive to trends

∑ ((Weight for period n)(Demand in period n))

WMA = ----------------------------------------------------------
∑Weights
 Qualitative Forecast Method (Time Series Methods)
3. Weighted Moving Average
∑ ((Weight for period n)(Demand in period n))
WMA = --------------------------------------------------------------------
∑Weights Month ACTUAL 3-month
weighted
z DEMAND
moving
Weights Applied Period (At) average
3 Last month 1 310
2 Two months ago 2 360
1 Three months ago 3 380
6 Sum of weights 4 415
3 x 380 + 2 x 360 + 1 x 310 5 450
6 6 470
7
 Qualitative Forecast Method (Time Series Methods)
3. Weighted Moving Average
∑ ((Weight for period n)(Demand in period n))
WMA = --------------------------------------------------------------------
∑Weights Month ACTUAL 3-month
weighted
z DEMAND
moving
Weights Applied Period (At) average
3 Last month 1 310
2 Two months ago 2 360
1 Three months ago 3 380
6 Sum of weights 4 415
3 x 380 + 2 x 360 + 1 x 310 5 450
6 6 470
7
 Qualitative Forecast Method (Time Series Methods)
3. Weighted Moving Average
∑ ((Weight for period n)(Demand in period n))
WMA = --------------------------------------------------------------------
∑Weights Weights Applied Period
Month
z ACTUAL
DEMAND (At)
3-month
weighted
3 Last month
moving
average 2 Two months ago

1 310 1 Three months ago

2 360 6 Sum of weights

3 380 361.67 (3 x 380) + (2 x 360) + (1 x 310)

6
4 415
1140 + 720 + 310
5 450
6
6 470 2,170
7 6
 Qualitative Forecast Method (Time Series Methods)
3. Weighted Moving Average
∑ ((Weight for period n)(Demand in period n))
WMA = --------------------------------------------------------------------
∑Weights Month ACTUAL 3-month
weighted
z DEMAND
moving
(At) average
(3 x 415) + (2 x 380) + (1 x 360)
6 1 310 -
1245 + 760 + 360 2 360 -
6 3 380 -
2,365 4 415 361.67
6
5 450 394.17
= 394.17
6 470 426.67
7 454.17
 Qualitative Forecast Method (Time Series Methods)
4. Exponential Smoothing
- This method uses a combination of the last actual demand and
the last forecast to produce the forecast for the next period.
z
New forecast = Last period’s forecast + α (Last period’s actual demand – Last period’s forecast)

Ft = Ft-1 + α (At-1 – Ft-1)

Where: Ft = new forecast

Ft -1 = previous period’s forecast
α = smoothing (or weighting) constant (0 ≤ α ≤ 1)
At-1 = previous period’s actual demand
 Qualitative Forecast Method (Time Series Methods)
4. Exponential Smoothing
- It is an easy method that enables forecasts to quickly react to
new trends or changes. A benefit to exponential smoothing is
that it does
z not require a large amount of historical data.
Exponential smoothing requires the use of a smoothing
coefficient called Alpha (α). The Alpha that is chosen will
determines how quickly the forecast responds to changes in
demand. It is also referred to as the Smoothing Factor.
 Qualitative Forecast Method (Time Series Methods)
4. Exponential Smoothing
Example:
In January, a car dealer predicted February demand for 142 Ford Mustangs.
z demand was 153 autos. Using a smoothing constant chosen
Actual February
by management of a = 0.20, the dealer wants to forecast March demand
using the exponential smoothing model.

New forecast (for March demand) = 142 +.2(153-142) = 142+2.2 = 144.2

Using just two pieces of data, the forecast and the actual demand, plus a
smoothing constant, we developed a forecast of 144 Ford Mustangs for
March.
 Qualitative Forecast Method (Time Series Methods)
4. Exponential Smoothing Ft = Ft-1 + α (At-1 – Ft-1)
Example: Ft = 361.67 + 0.20(415 – 361.67)
Ft = 361.67 + 0.20(53.33)
Period Actual Forecast Smoothing Smoothing Ft = 361.67 + 10.67
Ft= 372.34
Demandz Demand Constant Constant
a=0.2 a=0.5
Ft = 450 + 0.20(450 -394.17)
1 310 - Ft = 450 + 0.20( 55.83)
2 360 - Ft = 450 + 11.17
Ft =461.17
3 380 -
4 415 361.67 Ft = 470 + 0.20(470 - 426.67)
5 450 394.17 372.34 Ft = 470 + 0.20( 43.33)
Ft = 470 + 8.67
6 470 426.67 461.17
Ft = 478.67
7 454.17
 Qualitative Forecast Method (Time Series Methods)
4. Exponential Smoothing Ft = Ft-1 + α (At-1 – Ft-1)
Example: Ft = 361.67 + 0.50(415 – 361.67)
Ft = 361.67 + 0.50(53.33)
Period Actual Forecast Smoothing Smoothing Ft = 361.67 + 26.67
Ft= 388.34
Demandz Demand Constant Constant
a=0.2 a=0.5
Ft = 450 + 0.50(450 -394.17)
1 310 - Ft = 450 + 0.50( 55.83)
2 360 - Ft = 450 + 27.91
Ft =477.91
3 380 -
4 415 361.67 Ft = 470 + 0.50(470 - 426.67)
5 450 394.17 372.34 388.34 Ft = 470 + 0.50( 43.33)
Ft = 470 + 21.67
6 470 426.67 461.17 477.91
Ft = 491.67
7 454.17 478.67 491.67
 Qualitative Forecast Method (Time Series Methods)
5. Seasonal Index
- Many organizations produce goods whose demand is related to
the seasons, or changes in weather throughout the year. In
these cases,
z a seasonal index may be used to assist in the
calculation of a forecast.
- Using these calculated indices, we can forecast the demand for
next year based on the expected annual demand for the next
year.
 Qualitative Forecast Method (Time Series Methods)
5. Seasonal Index
Total Average Annual Demand
aAverage monthly demand = ------------------------------------------

z Period
Average monthly demand to past 3 years
bSeasonal index = -------------------------------------------------------

Average monthly demand

 Qualitative Forecast Method (Time Series Methods)
5. Seasonal Index
Example:

z
A Des Moines distributor of Sony laptop computers wants to
develop monthly indices for sales. Data from the past 3 years, by
month, are available
 Qualitative Forecast Method (Time Series Methods)
Demand
5. Seasonal Index Month Year
1
Year
2
Year 3 Ave.
yearly
Ave.
monthly
Seasonal
Indexb
Total yearly demand demand demanda
Ave. yearly Demand = --------------------------------- Jan. 80 85 105 90 94 .957(=90/94)
Period Feb. 70 85 85 80 94 .851(=80/94)
March 80 93 82 85 94 .904(=85/94)
Total Ave. Annual Demand April 90 95 115 100 94 1.064(=100/
Ave. Monthly Demand = ------------------------------ 94)
z Period May 113 125 131 123 94 1.309(=123/
94)
Ave. Monthly Demand to past 3 years June 110 115 120 115 94 1.223(=115/
94)
Seasonal Index = ----------------------------------------
July 100 102 113 105 94 .957(=90/94)
Average Monthly Demand
Aug. 88 102 110 100 94 .957(=90/94)
Sept. 85 90 95 90 94 .957(=90/94)
Oct. 77 78 85 80 94 .957(=90/94)
No. 75 82 83 80 94 .957(=90/94)
Dec. 82 78 80 80 94 .957(=90/94)
Total Average Annual 1,128
Demand
 Qualitative Forecast Method (Time Series Methods)
Demand
5. Seasonal Index Month Year Year 2 Year 3 Ave. Ave. Seasonal
1 yearly monthly Indexb
80 + 85+ 105 demand demanda
Average Yearly Demand = ----------------- = 90 Jan. 80 85 105 90 94 .957(=90/94)
3 Feb. 70 85 85 80 94 .851(=80/94)
March 80 93 82 85 94 .904(=85/94)
April 90 95 115 100 94 1.064(=100/9
1,128
4)
z = --------------- = 94
Average Monthly Demand May 113 125 131 123 94 1.309(=123/9
12 months 4)
June 110 115 120 115 94 1.223(=115/9
Seasonal Index = 90/94 = 0.957 4)
July 100 102 113 105 94 .957(=90/94)
Aug. 88 102 110 100 94 .957(=90/94)
Sept. 85 90 95 90 94 .957(=90/94)
Oct. 77 78 85 80 94 .957(=90/94)
No. 75 82 83 80 94 .957(=90/94)
Dec. 82 78 80 80 94 .957(=90/94)
Total Average Annual 1,128
Demand
 Qualitative Forecast Method (Time Series Methods)
5. Seasonal Index
If we expect the annual demand for computers to be 1,200 units next year,
we would use these seasonal indices to forecast the monthly demand as
follow Months Avg. Seasonal Index New Forecast
demand/month
z January 100 .957 95.6
February 100 .851 85.1
March 100 .904 90.4
April 100 1.064 106.4
May 100 1.309 130.9
June 100 1.223 122.3

❖Average demand = 1200 / 12 months = 100

❖New Forecast = Average demand x Seasonal Index
 Quantitative
z
Forecast
Associative Models
 Qualitative Forecast Method
(Associative Forecasting Methods)
Associative forecasting methods is different from time series
models because it considers other variables related to the
demand being predicted whereas the later only uses time to
z
predict demand.
Examples of other variables associate in demand or sales are:
1. Advertising expense,
2. Foot traffic,
3. Competitor’s data
4. Even the economic standing of a country
Associative Forecasting Methods (Regression analysis)
Linear Regression Formula 𝒚 = 𝒂 + 𝒃𝒙

Where y = value of the dependent variable

a =z y – axis intercept
b = slope of the regression line
X = the independent variables
Ʃ xy - n𝒙𝒚
b = -------------------
Ʃx2 -𝐧𝒙2
Associative Forecasting Methods (Regression analysis)
Linear Regression Formula 𝒚 = 𝒂 + 𝒃𝒙
Period Actual Level of
Calls Customer
Where:
(y) Satisfaction
(x) 1 = Disappointed
1 2.5 4.6 2 = Dissatisfied
2
z
2.7 4.56
3 = Neither Dissatisfied nor
3 3.4 4.03
satisfied
4 3 4.12
5 4 2.15
4 = satisfied
6 3.4 3.45 5 = delighted
7 4.2 1.88
8 4.7 2.38
9 3.8 3.48
10 4.8 2.23
Associative Forecasting Methods (Regression analysis)
Period Actual Level of
Calls Customer x2 xy
(y) Satisfaction
(x)
1 2.5 4.6 21.34 11.55
2 2.7 4.56 20.79 12.31
Ʃ xy - n𝒙𝒚 3 3.4 4.03 16.24 13.70
z 4 3 4.12 16.97 12.36
b = ---------------- 5 4 2.15 4.62 8.60
Ʃx2 -𝐧𝒙2 6 3.4 3.45 11.90 11.73
7 4.2 1.88 3.53 7.90
8 4.7 2.38 5.66 11.19
9 3.8 3.48 12.11 13.22
10 4.8 2.23 4.97 10.70
Ʃy Ʃx =32.90 118.16 113.26
=36.50
𝑦 − 3.65 𝑥=3.29
Associative Forecasting Methods (Regression analysis)
Ʃ xy - n𝒙𝒚 Period Actual Level of
b = ---------------- Calls Customer x2 xy
Ʃx2 -𝐧𝒙2 (y) Satisfaction
(x)
113.26 – (10)(3.29)(3.65)
b = ----------------------------- 1 2.5 4.6 21.34 11.55
118.16 – (10)(3.29)2 2 2.7 4.56 20.79 12.31
113.26 – (10)(12.01) 3 3.4 4.03 16.24 13.70
z
b= ---------------------- 4 3 4.12 16.97 12.36
118.16 – (10)(10.82) 5 4 2.15 4.62 8.60
113.26 – 120.10 6 3.4 3.45 11.90 11.73
b=----------------------- 7 4.2 1.88 3.53 7.90
118.16 – 108.20
8 4.7 2.38 5.66 11.19
-6.84
9 3.8 3.48 12.11 13.22
b=------
10 4.8 2.23 4.97 10.70
9.96
Ʃy Ʃx =32.90 118.16 113.26
b= -0.69 =36.50
𝑦 = 3.65 𝑥=3.29
Associative Forecasting Methods (Regression analysis)
Ʃ xy - n𝒙𝒚 Period Actual Level of
b = ---------------- Calls Customer x2 xy
Ʃx2 -𝐧𝒙2 (y) Satisfaction
(x)
113.26 – (10)(3.29)(3.65)
b = ----------------------------- 1 2.5 4.6 21.34 11.55
118.16 – (10)(3.29)2 2 2.7 4.56 20.79 12.31
113.26 – (10)(12.01) 3 3.4 4.03 16.24 13.70
z
b= ---------------------- 4 3 4.12 16.97 12.36
118.16 – (10)(10.82) 5 4 2.15 4.62 8.60
113.26 – 120.10 6 3.4 3.45 11.90 11.73
b=----------------------- 7 4.2 1.88 3.53 7.90
118.16 – 108.20
8 4.7 2.38 5.66 11.19
-6.84
9 3.8 3.48 12.11 13.22
b=------
10 4.8 2.23 4.97 10.70
9.96
Ʃy Ʃx =32.90 118.16 113.26
b= -0.69 =36.50
𝑦 = 3.65 𝑥=3.29
Associative Forecasting Methods (Regression analysis)
Ʃ xy - n𝒙𝒚 Period Actual Level of
b = ---------------- Calls Customer x2 xy
Ʃx2 -𝐧𝒙2 (y) Satisfaction
(x)
113.26 – (10)(3.29)(3.65)
b = ----------------------------- 1 2.5 4.6 21.34 11.55
118.16 – (10)(3.29)2 2 2.7 4.56 20.79 12.31
113.26 – (10)(12.01) 3 3.4 4.03 16.24 13.70
z
b= ---------------------- 4 3 4.12 16.97 12.36
118.16 – (10)(10.82) 5 4 2.15 4.62 8.60
113.26 – 120.10 6 3.4 3.45 11.90 11.73
b=----------------------- 7 4.2 1.88 3.53 7.90
118.16 – 108.20
8 4.7 2.38 5.66 11.19
-6.84
9 3.8 3.48 12.11 13.22
b=------
10 4.8 2.23 4.97 10.70
9.96
Ʃy Ʃx =32.90 118.16 113.26
b= -0.69 =36.50
𝑦 = 3.65 𝑥=3.29
Associative Forecasting Methods (Regression analysis)
Ʃ xy - n𝒙𝒚 Period Actual Level of
b = ---------------- Calls Customer x2 xy
Ʃx2 -𝐧𝒙2 (y) Satisfaction
(x)
113.26 – (10)(3.29)(3.65)
b = ----------------------------- 1 2.5 4.6 21.34 11.55
118.16 – (10)(3.29)2 2 2.7 4.56 20.79 12.31
113.26 – (10)(12.01) 3 3.4 4.03 16.24 13.70
z
b= ---------------------- 4 3 4.12 16.97 12.36
118.16 – (10)(10.82) 5 4 2.15 4.62 8.60
113.26 – 120.10 6 3.4 3.45 11.90 11.73
b=----------------------- 7 4.2 1.88 3.53 7.90
118.16 – 108.20
8 4.7 2.38 5.66 11.19
-6.84
9 3.8 3.48 12.11 13.22
b=------
10 4.8 2.23 4.97 10.70
9.96
Ʃy Ʃx =32.90 118.16 113.26
b= -0.69 =36.50
𝑦 = 3.65 𝑥=3.29
Associative Forecasting Methods (Regression analysis)
Ʃ xy - n𝒙𝒚 Period Actual Level of
b = ---------------- Calls Customer x2 xy
Ʃx2 -𝐧𝒙2 (y) Satisfaction
(x)
113.26 – (10)(3.29)(3.65)
b = ----------------------------- 1 2.5 4.6 21.34 11.55
118.16 – (10)(3.29)2 2 2.7 4.56 20.79 12.31
113.26 – (10)(12.01) 3 3.4 4.03 16.24 13.70
z
b= ---------------------- 4 3 4.12 16.97 12.36
118.16 – (10)(10.82) 5 4 2.15 4.62 8.60
113.26 – 120.10 6 3.4 3.45 11.90 11.73
b=----------------------- 7 4.2 1.88 3.53 7.90
118.16 – 108.20
8 4.7 2.38 5.66 11.19
-6.84
9 3.8 3.48 12.11 13.22
b=------
10 4.8 2.23 4.97 10.70
9.96
Ʃy Ʃx =32.90 118.16 113.26
b= -0.69 =36.50
𝑦 = 3.65 𝑥=3.29
Associative Forecasting Methods (Regression analysis)
Ʃ xy - n𝒙𝒚 Period Actual Level of
b = ---------------- Calls Customer x2 xy
Ʃx2 -𝐧𝒙2 (y) Satisfaction
(x)
113.26 – (10)(3.29)(3.65)
b = ----------------------------------------- 1 2.5 4.6 21.34 11.55
118.16 – (10)(3.29)2 2 2.7 4.56 20.79 12.31
3 3.4 4.03 16.24 13.70
113.26 –z (10)(12.01) 4 3 4.12 16.97 12.36
b= ------------------------------------------ 5 4 2.15 4.62 8.60
118.16 – (10)(10.82) 6 3.4 3.45 11.90 11.73
7 4.2 1.88 3.53 7.90
113.26 – 120.10
8 4.7 2.38 5.66 11.19
b= ------------------------------------------
9 3.8 3.48 12.11 13.22
118.16 – 108.20
10 4.8 2.23 4.97 10.70

-6.84 Ʃy Ʃx =32.90 118.16 113.26

b= --------------------------------- = 0.69 =36.50
9.96 𝑦 = 3.65 𝑥=3.29
Associative Forecasting Methods (Regression analysis)
Period Actual Level of
a= 𝒚 -b𝒙 Calls Customer x2 xy
(y) Satisfaction
(x)
a=3.65 – (-0.69)(3.29)
1 2.5 4.6 21.34 11.55
2 2.7 4.56 20.79 12.31
a= 3.65 –(-2.27) 3 3.4 4.03 16.24 13.70
z 4 3 4.12 16.97 12.36
a= 3.65 – (+2.27) 5 4 2.15 4.62 8.60
6 3.4 3.45 11.90 11.73
7 4.2 1.88 3.53 7.90
a= 3.65 + 2.27
8 4.7 2.38 5.66 11.19
9 3.8 3.48 12.11 13.22
a= -5.92 10 4.8 2.23 4.97 10.70
Ʃy Ʃx =32.90 118.16 113.26
=36.50
𝑦 − 3.65 𝑥=3.29
Associative Forecasting Methods (Regression analysis)
Period Actual Level of
Y = a + bx Calls Customer x2 xy
(y) Satisfaction
(x)
y=5.92 + (-0.69)(5)
1 2.5 4.6 21.34 11.55
2 2.7 4.56 20.79 12.31
y= 5.92 + (-3.44) 3 3.4 4.03 16.24 13.70
z 4 3 4.12 16.97 12.36
5 4 2.15 4.62 8.60

y= 2.48 6 3.4 3.45 11.90 11.73

7 4.2 1.88 3.53 7.90
8 4.7 2.38 5.66 11.19
9 3.8 3.48 12.11 13.22
10 4.8 2.23 4.97 10.70
Ʃy Ʃx =32.90 118.16 113.26
=36.50
𝑦 − 3.65 𝑥=3.29
Quantitative Forecast
z
Measurement of
Error
 Measurement of Error
1. Mean Absolute Deviation (MA)
2. Mean Squared Error (MSE)
3. Mean Absolute Percent Error (MAPE)
z
 Measurement of Error
1. Mean Absolute Deviation (MA)
2. Mean Squared Error (MSE)
3. Mean Absolute Percent Error (MAPE)
z

Forecast error/Deviation = Actual demand – Forecast Value

= At – Ft
 Measurement of Error
1. Mean Absolute Deviation (MA)
∑ [Actual – Forecast]
MAD = -------------------------------
n
2. Mean Squared Error (MSE)
z
∑ (Forecast errors)2
MSE = -------------------------------
n

3. Mean Absolute Percent Error (MAPE)

∑ absolute percentage error
MAPE = ----------------------------------------------
n
 Measurement of Error
Determine how reliable the forecast results based on the
following data using the following:
Period Actual Forecast
a. MAD
b. MSE
z
c. MAPE
1 310 315
2 365 375
3 395 390
4 415 405
5 450 435
6 465 480
 Measurement of Error
Mean Absolute Deviation (MAD):

To eliminate the problem of positive errors canceling negative

errors, a simple measure is one that looks at the absolute
value of the error (size of the deviation, regardless of sign).
z
When we disregard the sign and only consider the size of the
error, we refer to this deviation as the absolute deviation. If
we accumulate these absolute deviations over time and find
the average value of these absolute deviations, we refer to
this measure as the mean absolute deviation (MAD).
 Measurement of Error
a. Mean Absolute Deviation (MAD):

Period Actual Forecast Forecast Absolute

Error Deviation
At - Ft |At – Ft|
z
1 310 315 -5 5
2 365 375 -10 10
3 395 390 5 5
4 415 405 10 10
5 450 435 15 15
6 465 480 -15 15
 Measurement of Error
a. Mean Absolute Deviation (MAD):
Period Actual Forecast Forecast Absolute
∑ [Actual – Forecast] Error Deviation
MAD = --------------------------------- At - Ft |At – Ft|
n
1 310 315 -5 5
z 2 365 375 -10 10
3 395 390 5 5
4 415 405 10 10
5 450 435 15 15
6 465 480 -15 15
Total Absolute Deviation 60
Mean Absolute Deviation 60/6 =
10
 Measurement of Error
b. Mean Squared Error (MSE):

Another way to eliminate the problem of positive errors

canceling negative errors is to square the forecast error.
Regardlessz of whether the forecast error has a positive or
negative sign, the squared error will always have a positive
sign. If we accumulate these squared errors over time and find
the average value of these squared errors, we refer to this
measure as the mean squared error (MSE).
 Measurement of Error
b. Mean Squared Error (MSE)
Period Actual Forecast Forecast Squares Error
Error (At – Ft)2
At - Ft
1 310 315 -5 25
z2 365 375 -10 100
3 395 390 5 25
4 415 405 10 100
5 450 435 15 225
6 465 480 -15 225
Total Square Error 700
Mean Square Error 700/6 =
116.67
 Measurement of Error
c. Mean Absolute Percent Error (MAPE)

MAPE is computed as the average of the absolute difference

between the forecasted and actual values, expressed as a
percentage
z of the actual values.
 Measurement of Error
c. Mean Absolute Percent Error (MAPE)
Period Actual Forecast Forecast Error Absolute % Error
At - Ft 100(|At – Ft|/Actual)

1 310 315 -5 1.61%

2z 365 375 -10 2.74%
3 395 390 5 1.27%
4 415 405 10 2.41%
5 450 435 15 3.33%
6 465 480 -15 3.23%
Total Absolute % Error 14.59%
Mean Absolute % Error 14.59/6 = 2.43%
 Monitoring and
Controlling Forecast
z
 Tracking Signal
Tracking signal
➢ measurement of how well a forecast is predicting actual values.
➢ It is the ratio of cumulative forecast errors to mean absolute
deviation
➢ Positiveztracking signals indicate that demand is greater than
forecast.
➢ Negative signals mean that the demand is less than forecast.
➢ A good tracking signal has about as much positive error as it has
negative error.
➢ Good tracking signal has low values
➢ If forecasts are continually high or low, the forecast has a bias
error
 Tracking Signal

Formula:
Cumulative Error
zTracking Signal = ------------------------------
MAD
❖ The range of the tracking signal is a pair of the most
negative and most positive value
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake
forecast. Using the forecast and demand data for the past 6
quarters for cake sales we develop a tracking signal as follows:
z
Quarte Actual Forecas
r Deman t
d Deman
d
1 90 100
2 95 100
3 115 100
4 100 110
5 125 110
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake
forecast. Using the forecast and demand data for the past 6
quarters for cake sales we develop a tracking signal as follows:
Quarte z
Actual Forecas Error Cumulativ Absolute Cumulativ MAD Tracking Signal
r Deman t e Error Forecast e (Cum. Error/MAD)
d Deman Error Absolute
d Error
1 90 100
2 95 100
3 115 100
4 100 110
5 125 110
6 140 110
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake
forecast. Using the forecast and demand data for the past 6
quarters for cake sales we develop a tracking signal as follows:
Quarte z
Actual Forecas Error Cumulativ Absolute Cumulativ MAD Tracking Signal
r Deman t e Error Forecast e (Cum. Error/MAD)
d Deman Error Absolute
d Error
1 90 100 -10
2 95 100 -5
3 115 100 +15
4 100 110 -10
5 125 110 +15
6 140 110 +30
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake
forecast. Using the forecast and demand data for the past 6
quarters for cake sales we develop a tracking signal as follows:
Quarte z
Actual Forecas Error Cumulativ Absolute Cumulativ MAD Tracking Signal
r Deman t e Error Forecast e (Cum. Error/MAD)
d Deman Error Absolute
d Error
1 90 100 -10 -10
2 95 100 -5 -15
3 115 100 +15 0
4 100 110 -10 -10
5 125 110 +15 +5
6 140 110 +30 +35
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake
forecast. Using the forecast and demand data for the past 6
quarters for cake sales we develop a tracking signal as follows:
Quarte z
Actual Forecas Error Cumulativ Absolute Cumulativ MAD Tracking Signal
r Deman t e Error Forecast e (Cum. Error/MAD)
d Deman Error Absolute
d Error
1 90 100 -10 -10 10
2 95 100 -5 -15 5
3 115 100 +15 0 15
4 100 110 -10 -10 10
5 125 110 +15 +5 15
6 140 110 +30 +35 30
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake
forecast. Using the forecast and demand data for the past 6
quarters for cake sales we develop a tracking signal as follows:
Quarte z
Actual Forecas Error Cumulativ Absolute Cumulativ MAD Tracking Signal
r Deman t e Error Forecast e (Cum. Error/MAD)
d Deman Error Absolute
d Error
1 90 100 -10 -10 10 10
2 95 100 -5 -15 5 15
3 115 100 +15 0 15 0
4 100 110 -10 -10 10 10
5 125 110 +15 +5 15 5
6 140 110 +30 +35 30 35
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake
forecast. Using the forecast and demand data for the past 6
quarters for cake sales we develop a tracking signal as follows:
Quarte z
Actual Forecas Error Cumulativ Absolute Cumulativ MAD Tracking Signal
r Deman t e Error Forecast e (Cum. Error/MAD)
d Deman Error Absolute
d Error
1 90 100 -10 -10 10 10 10
2 95 100 -5 -15 5 15 7.5
3 115 100 +15 0 15 30 10
4 100 110 -10 -10 10 40 10
5 125 110 +15 +5 15 55 11
6 140 110 +30 +35 30 85 14.2
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake
forecast. Using the forecast and demand data for the past 6
quarters for cake sales we develop a tracking signal as follows:
Quarte z
Actual Forecas Error Cumulativ Absolute Cumulativ MAD Tracking Signal
r Deman t e Error Forecast e (Cum. Error/MAD)
d Deman Error Absolute
d Error
1 90 100 -10 -10 10 10 10 -10/10 = -1
2 95 100 -5 -15 5 15 7.5 -15/7.5 = -2
3 115 100 +15 0 15 30 10 0/10 = 0
4 100 110 -10 -10 10 40 10 -10/10 = -1
5 125 110 +15 +5 15 55 11 +5/11 = +0.5
6 140 110 +30 +35 30 85 14.2 +35/14.2 = +2.5
 Tracking Signal
Example :
Carlison’s Bakery wants to evaluate performance of cake forecast.
Using the forecast and demand data for the past 6 quarters for cake sales
we develop a tracking signal as follows:
Quarte Actual Forecas Error Cumulativ Absolute Cumulativ MAD Tracking Signal
r Deman t e Error Forecast e (Cum. Error/MAD)
d Deman Error Absolute
z d Error
1 90 100 -10 -10 10 10 10 -10/10 = -1
2 95 100 -5 -15 5 15 7.5 -15/7.5 = -2
3 115 100 +15 0 15 30 10 0/10 = 0
4 100 110 -10 -10 10 40 10 -10/10 = -1
5 125 110 +15 +5 15 55 11 +5/11 = +0.5
6 140 110 +30 +35 30 85 14.2 +35/14.2 = +2.5
Insight – Because the tracking signal drifted from -2 MAD to +2 MAD
(between 1.6 and 2.0 standard deviations), we can conclude that it is within
 Thank you!
z