BASIC IDEA: End products are typically comprised of several components and, as a result, sales are interrelated. Input-output analysis is used to identify the linkages between the products and thus can be useful in developing better forecasts.
PROCEDURE:
Identify the interrelationships
Quantify the interrelationships
Construct an input-output table that represents these interrelationships. The table gives the total requirements (inputs), both direct and indirect. Thus, this analysis is particularly useful for uncovering situations where one product indirectly supplies another.
EXAMPLE: The following chart outlines a simple Input/Output matrix that details the relationship between PC's, laser printers, and DRAMs.
Inputs |
|||
Outputs |
PC |
Laser Printers |
DRAMS |
PC |
|||
Laser Printers |
.155 |
||
DRAMs (MB) |
1.53
|
2.25 |
Thus, there are about .155 laser printers per PC, 1.53MB of DRAM per PC, and 2.25 MB of DRAM per laser printer. To forecast sales for DRAMs, you could use the sales from laser printers and PCs as a basis. However, recall that these laser printer and PC sales numbers would be forecasts as well.
COMMENTS:
In general, you need many data points to get a good estimate of the Input-Output ratio. Without many past data points the resulting estimate becomes more of a subjective estimate than one which is highly rigorous. Nevertheless, if these ratios are treated as part of an overall judgmental approach, they can be used to provide additional insights to the forecast.
There are strong assumptions associated with input-output analysis:
Relationships are the same for all levels of production. Tenuous if scale economies are present.
Relationships hold true over time. Tenuous in rapidly changing markets.