Outline of Presentation

• What is Forecasting?
• Forecasting Time Horizons
Demand Management
• Types of Forecasts
Forecasting Methods
• Forecasting Approaches
– Qualitative Methods
– Quantitative Methods

What is Forecasting? Forecasting Time Horizons

• Forecasting is the art and science of • Short-Range Forecasts

predicting future events. It may involve – This forecast has a time span of up to 1
– Taking historical data and projecting it into year but is generally less than 3
the future using mathematical models months.
– Subjective or intuitive prediction – It is used for planning purchasing, job
– A combination of the aforementioned scheduling, workforce levels, job
statements assignments, and production levels.

1
Forecasting Time Horizons Forecast Time Horizons

• Medium-Range Forecasts • Long-Range Forecasts

– A medium-range or intermediate forecast – Generally 3 years or more in time span
generally spans from 3 months to 3 – Long-range forecasts are used in planning
years. for new products, capital expenditures,
– It is useful in sales planning, production facility location or expansion, and research
planning and budgeting, cash budgeting, and development
and analysis of various operating plans.

Types of Forecasts Types of Forecasts

• Economic Forecasts • Technological Forecasts

– Address the business cycle by predicting – It is concerned with rates of technological
inflation rates, money supplies, housing progress, which can result in the birth of
starts, and other planning indicators. exciting new products, requiring new plants
and equipment.

2
Types of Forecasts Forecasting Approaches

• Demand Forecasts • Qualitative Methods

– These are projections of demand for – Jury of Executive Opinion
company’s products or services. – Delphi Method
– These forecasts, also called sales – Sales Force Composite
forecasts, drive a company’s production, – Consumer Market Surveys
capacity, and scheduling systems and
serve as inputs to financial, marketing, and
personnel planning.

Forecasting Approaches Qualitative Methods

• Quantitative Methods • Jury of Executive Opinion

– Naive Approach – The opinions of a group of high-level
– Moving Average experts or managers, often in
– Exponential Smoothing combination with statistical models, are
pooled to arrive at a group estimate of
– Trend Projection demand.
– Linear Progression

3
Qualitative Methods Qualitative Methods
• Delphi Method • Sales Force Composite
– Decision Makers: group of 5 to 10 experts – Each salesperson estimates what sales
who will make the actual forecast will be in his or her region. After
– Staff Personnel: assist decision makers by reviewing the data, they combine at the
preparing, distributing, collecting, and district and national levels to reach an
summarizing a series of questionnaires overall forecast.
and survey results.
– Respondents: group of people, often
located in different places, whose judgment
are valued.

Qualitative Methods Quantitative Methods

• Consumer Market Survey • Naive Approach

– This method solicits input from – The simplest way to forecast is to assume
customers or potential customers that demand in the next period will be
regarding future purchasing plans. equal to demand in the most recent
period.
Period Actual Demand Forecast
January 45 -
February 60 45
March 72 60
April 58 72
May 40 58
June 40

4
Quantitative Methods Quantitative Methods

• Moving Averages • Example: Moving Averages

– It uses a number of historical actual data – Donna’s Garden Supply wants a 3-month
values to generate a forecast. It assumes moving average forecast, including a
that market demands will stay fairly steady forecast for next January, for shed sales.
over time. Actual Shed Actual Shed
Month Month
Sales Sales
January 10 July 26
Moving Σ Demand in previous n periods February 12 August 30
= ____________________________
Average March 13 September 28
n
April 16 October 18
May 19 November 16
June 23 December 14

Quantitative Methods Quantitative Methods

• Weighted Moving Averages • Example: Weighted Moving Averages

Period Weights Applied
– When a detectable trend or pattern is
Last month 3
present, weights can be used to place Two months ago 2
more emphasis on recent values. Three months ago 1
Sum of weights 6

Actual Shed Actual Shed

Moving = Σ_______________________________
(Weight for period n)(Demand in period n) Month
Sales
Month
Sales
Average Σ Weights January 10 July 26
February 12 August 30
March 13 September 28
April 16 October 18
May 19 November 16
June 23 December 14

5
Quantitative Methods Quantitative Methods

• Exponential Smoothing • Example: Exponential Smoothing

– The latest estimate of demand is equal to
the old estimate adjusted by a fraction of Quarter Actual Tonnage Forecast at 0.5
the difference between the last period’s 1 180 175
2 168
actual demand and the old estimate.
3 159

F F t t 1
( At 1 F t 1) 4 ?

Ft = new forecast
Ft-1 = previous period’s forecast
α = smoothing (or weighting) constant (0 ≤ α ≤ 1)
At-1 = previous period’s actual demand

Quantitative Methods Quantitative Methods

• Exponential Smoothing w/ Trend • Exponential Smoothing w/ Trend

Adjustment Adjustment

Forecast including Trend (FITt) = FITt = Ft + Tt

Exponentially smoothed forecast (Ft) +
Exponentially smoothed trend (Tt) Ft = α(At-1) + (1- α)(Ft-1 + Tt-1)

FITt = Ft + Tt Tt = β(Ft – Ft-1) + (1- β)Tt-1

6
Quantitative Methods Quantitative Methods

• Example: Exponential Smoothing w/ • Exponential Smoothing w/ Trend

Trend Adjustment Adjustment
– A review of past sales, as shown on the Month (t) Actual Demand (At)
next slide, indicates that an increasing 1 12
trend is present. Smoothing constants are 2 17
assigned the values of α =0.2 and β =0.4. 3 20
The firm assumes the initial forecast for 4 ?

month 1(F1) was 11 units and the trend

over that period (T1) was 2 units.

Quantitative Methods Quantitative Methods

• Trend Projections • General equation:

– This technique fits a trend line to a series
of historical data points and then projects ŷ = a + bx
the line into the future for medium to long
range forecasts. ŷ is the computed value of the variable to be predicted
– This method applies the concept of least- a = y-axis intercept
b = slope of the regression line (rate of change in ŷ)
squares method. x = is the independent variable (time)

7
Quantitative Methods Quantitative Methods

• How to solve for a and b: • Example: Trend Projections

– The demand for electric power at N.Y.
xy nx y Edison over the period 2003 to 2009 is
b 2 shown in the following table, in MW. The
2
x nx firm wants to forecast 2010 demand by
fitting a straight-line trend to these data.

a y bx

Quantitative Methods Quantitative Methods

• Example: Trend Projections • Linear Regression

Year
Electrical Power Demand, – Least-Squares Method can still be used to
in megawatts
perform the linear-regression analysis. But
2003 74
now the independent variable, x, need no
2004 79
2005 80
longer be time.
2006 90
2007 105
2008 142
2009 122

8
Quantitative Methods Quantitative Methods

• Example: Linear Regression • Example: Linear Regression

– Nodel Construction Company wants to – If the local chamber of commerce predicts
establish a mathematical relationship to that the West Bloomfield area payroll will
help predict sales. be $6 billion next year, what will be the
Nodel’s Sales Area Payroll
estimated sales for Nodel in the coming
(in $ millions), y (in $ billions), x year?
2.0 1
3.0 3
2.5 4
2.0 2
2.0 1
3.5 7

Demand Management

Reference:
Heizer, J., & Render, B. (2011).
Operations Management, 10th edition.