Forecasting - Chapter 18

Question 1

What is forecasting?

Answer 1

Predicting future demand based on past and present data - vital function that affects management decisions in different departments.

in OM: interested in forecasting future demand.

Question 2

Components of Demand:

Average Demand

Answer 2

this is the baseline demand the product or market is expected to have

Question 3

Trend

Answer 3

often long-term due to changes in population, technology and culture

• Linear Trend
• S-Curve Trend
• Asymptotic Trend
• Exponential Trend

Question 4

Seasonal Element

Answer 4

seasonality that repeats every 1-year period

Question 5

Random variation

Answer 5

purely random, unexplained variation

Question 6

Qualitative forecasting

Answer 6

• Used when situation is vague and little data exist
• New products and new market
• Involves intuition and expert knowledge
• Often use soft information such as personal opinions

Question 7

Quantitative forecasting

Answer 7

• Used when situation is stable and historical data exist
• Existing products and existing market
• Involves mathematical & statistical techniques
• These techniques rely on hard data (Historical sales data)

Question 8

Methods for qualitative forecasting

Answer 8

use subjective opinions from consumers, sales staffs, managers, executives, and experts.

-

Question 9

2. Market research in forecasting:

Answer 9

Uses consumer surveys (consumer opinions) and interviews to gather demand insights.

• Surveys are easy and low-cost (especially online surveys). However, proper execution needs resources and expertise.
• Firms often hire specialized companies, which is costly, but interviews provide in-depth information.

Question 10

3. Sales force opinions:

Answer 10

builds forecasts from the bottom up using input from sales or customer service staff who interact directly with customers.

• assumes those closest to the customer know best, but forecasts may be overly optimistic.

Question 11

4. Panel Forecast:

Answer 11

panel of people from a variety of positions that can develop a more reliable forecast.
- forecasts are developed through open discussion to improve reliability.

Question 12

Cons:

Answer 12

Lower-level employees might feel intimidated by higher levels of management

Question 13

Executive judgement (Jury of Executive opinion)

Answer 13

involves only upper-level managers who make quick decisions to make forecasts.

Question 14

5. Delphi Method:

Answer 14

is an iterative process to achieve a consensus on a forecast.

• usually achieves satisfactory results in 3-4 rounds.

Question 15

Step-by-step procedure for Delphi method:

Answer 15

1. Select experts (5-10 participants)
2. Create and distribute questionnaire
3. Collect anonymous responses
4. Summarize and redistribute results
5. Refine forecasts and repeat until consensus is reached
6. Distribute final results to all participants

Question 16

Limitations:

Answer 16

Participant quality
Strong opinions
Anonymity

Question 17

Methods of quantitative forecasting:

Answer 17

1. Time-series models
   - based on historical data, idea relates back to how past demand data can be used to predict future demand.
   Example: Moving Average, Weighted Moving Average, and Exponential Smoothing
2. Casual models
   - based on a predictor or a set of predictors other than the past demand information
   Example: regression models

Question 18

Naive forecast

Answer 18

uses a single previous value as the basis for a forecast
- assumes demand will stay the same as the most recent period.

Used in very stable markets with little fluctuation.

Question 19

Simple Moving Average (SMA)

Answer 19

uses the avg of a fixed number of past periods

• equal weighting for all periods

to calc this: you need to get however many months they say like 3 month SMA and from the month your trying to find for you take those last three months and find the avg.

Question 20

Weighted Moving Average (WMA)

Answer 20

unequal weighting of prior time periods, but must be = to one

–> choice is arbitrary but expert knowledge is helpful - more recent periods usually have higher weights (%) than older periods.

Question 21

Exponential Smoothing

Answer 21

weighted avg method that includes all past data and observations in the forecasting calculation

• more recent results are weighted more heavily - more significant - based on smoothing constant (ranges from 0-1)

Question 22

Formula E.S:

Answer 22

𝐹𝑡 = 𝐹𝑡−1 + 𝛼 (𝐴𝑡−1 − 𝐹𝑡−1)
– Ft = forecast demand for period t
– Ft-1=forecast demand for the prior period t-1
– At-1 = actual demand in the prior period t-1
– a = The smoothing constant alpha

Question 23

Is it possible to have a 100% accurate forecast?

Answer 23

No, all forecasts have some level of error. The goal is to minimize forecast error as much as possible.

Question 24

Formula forecast error:

Answer 24

Error = Actual – Forecast => 𝐴𝑡 − 𝐹𝑡
- (t= time period)

• can calculate the forecast error for past sales or only after the sales are realized.

Question 25

When and how can forecast error's be measured?

Answer 25

Forecast error can only be calculated after actual values are known. - For one-period forecasts: use At - Ft directly. - For multiple periods: further analysis is needed using error metrics.

Question 26

Error Metrics

Answer 26

1. Mean absolute deviation MAD 2. Mean square error MSE 3. Mean absolute percent error MAPE Traits: - MAD and MSE are commonly used - Larger values indicate forecasts are LESS accurate - MAPE is useful when you compare forecasts of different products with different demand

Question 27

Choosing a forecasting technique:

Answer 27

1. use the forecasting model that minimizes your avg forecasting error 2. cost-accuracy trade-off 3. good strategy: use various methods for a comprehensive view

Question 28

Collaborative planning, forecasting & replenishment (CPFR):

Answer 28

a collaborative process used to coordinate the efforts of a supply chain. Includes: demand forecasting, production and purchasing, and inventory replenishment

Question 29

Who is involved in CPFR:

Answer 29

integrates all members of supply chain. Example: manufacturers, distributors, and retailers - depends upon the exchange of internal information to provide a more reliable view of demand.