Forecasting Overview
Meijer Foods appears to use models from each of the four types of forecasting techniques. Some are stand-alone, others range between partial to totally automated.
When a customer purchases a product (at a cash register), a series of events occurs within the computer environment within minutes. The outcome is a proposed purchase order. For most product, one purchase, relates to one item ordered. Products with quick turn-around cause the shelf to empty faster. Therefore, the faster product is purchase off the shelf, the faster a purchase order will be issued by the system. All this is accomplished to insure product is always on shelves and hangers. A key feature of this system is the fact that if a product is not on the shelf, and not in a trailer backed up to the loading dock, it is not at the store. Most of the forecasting models relate to items not in stock at a store. Once a product is within a store, the system is automatic. The challenge then becomes, stopping automatic reorder of seasonal products. Part of that challenge is resolved by making reorders only to a pre-stocked warehouse. Once the warehouse is empty, the ability to reorder has ended.
Specific examples of the four types of forecasting techniques follow in the next paragraphs. Most of the models are used within the organization.
Qualitative Techniques
Chapter 12 (Forecasting) of the course material states: “Qualitative techniques are subjective or judgmental and are based on estimates and opinions” (p. 467). Some of the qualitative models are grass roots, market research, panel consensus, and historical analogy. Examples of these are:
Grass Roots A new product is proposed within the deli department.
Store department heads will be surveyed for their recommendations.
Market research The infrequent hostess on the floor within the grocery
department providing small quantities of sample product.
Panel Consensus A new product is introduced without a history. A
panel within the purchasing department would develop a
consensus.
Historical Analogy A new soup from Campbell’s will use a history from another Campbell’s soup product.
Time Series Analysis
Chapter 12 (Forecasting) of the course material states that time series analysis is “based on the idea that data relating to pass demand can be used to predict future demand” (p. 467). Some of the common time series analysis models are simple moving average, weighted moving average, exponential smoothing, regression analysis, and trend projections. Examples of these are:
Simple moving average Fish may be forecasted using this model. The
system would have several years of history to build
the average
Weighted moving average Turkey may be forecasted using this model. The
system would have several years of history to build
the average
Exponential smoothing Special order cuts of meat with the meat department may use this system to forecast basic stock
Regression analysis Prior to the automation, this would have been used
for the forecasting of bread
Trend projections. Before the automation, this would have been used
for the forecasting of paper, card and book products
Casual Forecasting
The course material within chapter 12 (Forecasting) states that casual forecasting, “assumes that demand is related to some underlying factor or factors in the environment” (p. 467). Common casual forecasting models are regression analysis, Econometric models, input/output models, and leading indicators. Examples of these are:
Regression analysis Before automation, ice cream forecasting could use
this model
Econometric models Before automation, home building material may
have used this model
Input/output models Before automation, computer component items
may have been forecast using this model.
Leading indicators. High-test gas may be forecast by the purchase of
regular gas.
Simulation models
Chapter 12 (Forecasting) of the course material states that simulation models are “dynamic models, usually computer-based that allow the forecaster to make assumptions about the internal variables and extreme environment in the model” (p. 468). No experience is available to determine if any simulation models were used.
Reference
University of Phoenix (2003). Operations Management for Completive Advantage, Tenth Edition [University of Phoenix Custom Edition e-text]. McGraw Hill Companies. Retrieved January 6, 2004, from University of Phoenix, Resource, MGT 554 - Operations Management Web Site https://mycampus.phoenix.edu/secure/ resource/resource.asp
