SYSPRO's Inventory Forecasting module is the first of three modules within the Inventory Optimization suite. The full suite of modules comprise:

Before inventory levels can be optimized, it is essential that sales forecasts are as accurate and reliable as possible. The Forecasting Module addresses this requirement by providing the tools to achieve this.

Given that the forecasting algorithms base their calculations exclusively on sales history, it is important that the sales history (as held in the forecasting system) is a true reflection of the sales demand made on the company's inventory (i.e. it should include any sales lost because there was no inventory available to satisfy a demand).

The Sales Orders module outputs invoices, debit notes, credit notes and dispatch notes to the forecasting sales history. In addition, the Initial History Creation program enables you to build the forecasting sales history using inventory movements as its source.

There is also a business object that can be used to take on forecasting sales history. This would be useful to companies that have recently migrated to SYSPRO.

The system provides optional automatically-generated 'outliers adjustment entries' to help ensure that abnormal historical sales patterns (that are not likely to re-occur) do not affect forecasts of future sales.

Finally, the Demand History Maintenance program presents the sales history (in both tabular and graphical format) and allows the user to create manual adjustment entries.

Faced with the daunting task of having to sort out the sales history and forecasts of a company's entire inventory, users might wonder where they should focus their attention. The Pareto Analysis program provides insight to which items are 'most important' to the company.

The analysis can be run on all inventory items or sub-sets as defined by the available options. The selected options can be saved to a 'forecasting set' for later use in Pareto Analysis and other forecasting module programs. Having analyzed their inventory using Pareto Analysis, the user can choose to manually forecast all 'A class' items while using the Batch Forecasting program to forecast the remainder.

Since demand history can be volatile, Pareto classifications may change and we recommend that you run Pareto Analysis at least once a month.

It is envisaged that the Batch Forecasting program will be used to forecast the bulk of a company's items. Typically, the items suitable for batch forecasting will have been identified using Pareto Analysis and the same 'forecasting set' of options used to run Batch Forecasting (perhaps excluding 'A class' items). The guiding principle here would be to automate the ordinary and focus on the exceptions.

The Forecast Comparison report (used to compare draft forecasts to current forecasts as a check for reasonableness) and the Approve Draft Forecasts program (which updates the current forecasts) both also allow the use of run-time options saved against 'forecasting sets'. The implementation of 'forecasting sets' enables users to have pre-defined groups of items against which the main forecasting module functions can be performed.

Of course, there is also the Manual Forecasting program that provides the user with an interactive forecasting process and shows (in both tabular and graphical format) the sales history, current forecasts, draft forecasts and suggested forecasts. Here, the user can compare the results of the various algorithms available and tweak the algorithms to suit the individual inventory items.

A Forecasting Measures tab allows you to assess fit errors for the algorithms selected.