Forecast Accuracy

Inventory Planning > Inventory Forecasting > Reports > Forecast Accuracy

Forecast Accuracy

You use this program to track the performance of the forecast for individual items or a group of items. This enables you to measure and compare your historical forecasts against your actual sales movements, including lost sales.

This report calculates the same information as the Forecast Accuracy Review program, so you do not have to run the Forecast Accuracy Review first.

The report should be run as part of the monthly forecast cycle to validate the forecast and to identify items with high forecast error. It can also be run at any time for all items or a group of items. Items with high forecast error or high forecast error value may require that their forecast be manually updated to reduce the error.

Forecast accuracy is an essential measure of how well the forecast is performing. It uses the actual error achieved (forecast - actual) over a period of months into the past selected at run time. It is therefore essential to store the forecasts made over a period of time to ensure that they can be compared to the actual sales achieved.

Report Options

Field

Description

Selection

Use selection set

Select this to run the report using a selection set. This option is selected by default and you indicate the selection set to use in the Selection set field below.

When you deselect this option, all the other options are enabled and you can indicate the selections you require.

Selection set

Indicate the selection set you want to report on.

Run date

Enter the report run date to use.

This run date enables you to run the report using forecast snapshots, which allows you to see the change in accuracy over time.

Options

Return period details

Select this to list the values within the Months to compare indicated.

Values

Indicate the values you want to print on the report.

One of the following can be selected:

Cost values
Sales values
Gross profit values (GP values)
Quantities

Forecast accuracy

These options are only available if you did not select to run the report using a selection set.

Calendar to use

Indicate the calendar to use to calculate the periods against which forecast accuracy is calculated.

Months to compare

Enter the number of months into the past to use to compare sales history against forecasts to check the forecast accuracy.

In the calculation, the maximum number of months will be less than or equal to the months of history available.

Comparison type

Select either period on period or moving average comparisons.

Option

Description

Period on period

Select this to perform a direct comparison between the forecast and sales periods.

Each month in the selected window into the past is compared to the forecast for the corresponding period.

Moving average

Select this to perform the comparison using moving averages.

The month average is calculated for both the actuals and the historic forecasts and compared.

A one-sided moth average is used at the beginning and end of the range and a two-sided month average is used where this is possible.

This option is useful when there is slow or erratic movement of a item.

The number of Months to compare define the range of the moving average calculation.

Moving average months

Enter the number of months to use to calculate a moving average for sales and forecasts.

Forecast to use

Select the forecast to compare to the sales.

A key consideration is which forecast to use when calculating accuracy for a particular month.

There may have been several forecasts made for the particular period (known as forecast shots). In general, the forecast accuracy should be based on a forecast that could influence the future.

Any forecasts made inside the lead time should not be used, as such forecasts may artificially improve the accuracy. The report addresses this by allowing for several options in calculating the accuracy.

Option	Description
Last	This is the last snapshot taken for a period.
Last outside lead time	This is the last snapshot taken outside of the lead time for the item.
Average	This will take an average of the snapshots that apply to a period.
Average outside of lead time	This will take an average of snapshots created for a period outside of lead time.

Snapshot months

Indicate how many months of forecasts you want to use to calculate the average forecast.

Return accuracy levels

Select the category or level of forecast accuracy you want to print on the report. This is the sequence in which the report is generated.

SKU accuracy levels lists the values against each individual stock item included in the report. Stock items are listed multiple times if they are stocked in more than one warehouse.

Consolidate accuracy levels lists the values against the stock items included in the report, but each item is listed only once (i.e. the values for the stock item in each warehouse are consolidated).

Pareto analysis must be run for entries to be accurately listed on Pareto accuracy levels (see Pareto Analysis).

Selection criteria

Selection	Description
Warehouse selection	Indicate the warehouse(s) for which you want to generate the report.
Stock code selection	Indicate the stock code(s) for which you want to generate the report.
Supplier selection	Indicate the supplier(s) for which you want to generate the report.
Product class selection	Indicate the product class(es) for which you want to generate the report.
Buyer selection	Indicate the buyer(s) for which you want to generate the report.
Planner selection	Indicate the planner(s) for which you want to generate the report.

Include

These options are only available if you are not using a selection set.

Field	Description
Part type	Indicate the part types to include in the report. Phantom, Planning Bill and Notional Part types have no bearing in the IO modules and are therefore not available here.
Pareto classification	This enables you to include/exclude items based on their Pareto classification. You use the Pareto Analysis program to analyse and classify inventory items. At least one of these classifications must be selected.
Analysis type	Indicate what the Pareto classification, for the inclusion of stock items, must be based on. This can be Sales, Current forecasts or Draft forecasts. Once you have selected an analysis type, you can select the Pareto level at which the analysis must be performed. For example, if you select Sales and Stock code, then the Pareto classification of items used for inclusion purposes will be retrieved based on sales at the stock code level.
Sales	Select this to perform the analysis based on sales.
Current forecasts	Select this to perform the analysis based on current forecasts. This allows you to compare expected contribution with past contribution.
Draft forecasts	Select this to perform the analysis based on draft forecasts.
Pareto level	Indicate the level at which the Pareto analysis must be performed.
Other	Against each of the following options, you can select: Yes to include the items in the report. No to exclude the items from the report. Only to only generate the report for those items.
Manual forecast items	Indicate how manually forecasted items must be handled for the report.
Replenishment items	Indicate how replenishment items must be handled for the report.
MPS items	Indicate how MPS items must be handled for the report.
Make to order items	Indicate how make to order items must be handled for the report.

Output Options

These options enable you to apply a theme to the report and to define multiple output destinations for the report once it has been compiled (SRS Output Options).

Report Details

The report is generated in the sequence you selected at the Return accuracy levels option.
An entry of {Unclassified} indicates that the information is not saved in the Inventory Master file/table. For example all selected stock items that do not have the buyer defined against them are listed in the unclassified section for Buyers on the report.

The following information is included in the report:

Column	Description
Forecast	This indicates the forecast value for the forecast used.
Sales	This indicates the actual sales value used to compare against the forecast value.
Hits	This indicates the number of actual sales.
CFE	Cumulative forecast error. This is the cumulative difference between the forecast and the actual sale(s) for the period under review. A positive number indicates a forecast that is too high for a period and a negative number indicates a forecast which is too low for the period.
MAD	Mean Absolute Deviation. The is the average absolute deviation from the mean (i.e. the average error, ignoring the sign of the error).
MSD	Mean square deviation. This is the sum of the squared forecast errors for each observation divided by the number of observations. It is an alternative to the Mean Absolute Deviation except that, because the errors are squared, more weight is placed on larger errors.
MAPE	Mean Absolute Percentage Error. This is calculated as the average of the sum of all the absolute percentage errors for the data set. The absolute values of the percentages are summed and the average is computed. The difference between actual value and the forecast value is divided by the actual value. The absolute value of this calculation is summed for every forecast point in time and divided again by the number of fitted points.
Tracking signal	This is calculated as the algebraic sum of forecast errors divided by the Mean Absolute Deviation (MAD). It indicates whether the forecast average is keeping pace with any genuine upward or downward changes in demand.