Demand Forecasting - PowerPoint PPT Presentation

1 / 23
About This Presentation
Title:

Demand Forecasting

Description:

... views of key executives to obtain a sounder sales forecast than might be made by ... In some companies, each salesperson estimates the future sales in his ... – PowerPoint PPT presentation

Number of Views:3580
Avg rating:3.0/5.0
Slides: 24
Provided by: henr80
Category:

less

Transcript and Presenter's Notes

Title: Demand Forecasting


1
Demand Forecasting
  • Henry C. Co
  • Technology and Operations Management,
  • California Polytechnic and State University

2
Types of Models
  • Qualitative based on experience, judgment,
    knowledge
  • Quantitative based on data, statistics.

3
Qualitative Forecasts
  • Executive opinions
  • Combines views of key executives to obtain a
    sounder sales forecast than might be made by a
    single estimator.
  • Sales force composite
  • Obtains the combined views of members of the
    sales force as to the future sales outlook. In
    some companies, each salesperson estimates the
    future sales in his or her territory.
  • To ensure realistic estimates, successive
    management levels are likely to do careful views.

4
Qualitative Forecasts
  • Consumer surveys
  • Involves asking product users about the
    quantities they expect to buy in the forecast
    period. By combining user responses, the
    interviewing firm can estimate total demand for
    the product (or service), and then determine the
    portion of that demand that it expects to fill.
  • Outside opinion
  • Opinions of managers and staff
  • Delphi technique

5
Qualitative Forecasts
  • Naive Methods eye-balling the numbers
  • Formal Methods systematically reduce
    forecasting errors
  • Time series models (e.g. exponential smoothing)
  • Causal models (e.g. regression).

6
Naive Forecasts
Uh, give me a minute.... We sold 250 wheels
last week.... Now, next week we should sell....
7
Time Series Forecasts
Seasonal variation
Trend
Level
Assumptions There is information about the
past This information can be quantified in the
form of data The pattern of the past will
continue into the future.
8
(No Transcript)
9
Using MS Excel
10
Select the range A1B22. Click Insert, select
charts (scatter).
11
Move the cursor to any point on the graph, and
right-click. Choose Add Trendline.
12
Move the cursor to any point on the graph, and
right-click. Choose Add Trendline. For example,
select Exponential. Forecast Forward 34 periods
(through 1989). Display Equation and R-squared
value on chart
13
(No Transcript)
14
Use the trend-line formula for post-diction y
7E-59e0.0753x. C2 7E-59EXP(0.0753A2) Copy
and Paste.
15
Select the range A1B22. Click Insert, select
charts (scatter). Modify chart, legend, etc.
16
Try other trend-linesPolynomial, etc.
  • The easiest way to do this is to make a copy of
    the worksheet for the Exponential trend-line.
    Move the cursor to the trend-line, right-click,
    and choose a different trend-line.

17
Polynomial Trend-Line
18
Moving Average
19
Forecast Errors
  • MAD or MSE?

20
  • Mean Squared Error (MSE)
  • Measures the accuracy of the forecasts
  • sum of squares of the forecast errors, divided
    by n-1. Here, n is the number of squared errors
    summed.
  • Mean Absolute Deviation (MAD)
  • Also measures of forecast accuracy.
  • average of the absolute value of forecast
    errors.

21
  • These two measures of forecast errors are usually
    consistent in the sense that if one forecasting
    model yields a higher MAD, it will also result in
    a higher MSE.
  • But there are exceptions

22
Past Performance of Two Models
23
Which model would you use?A or B?
Write a Comment
User Comments (0)
About PowerShow.com