Can I enter a forecast to drive MRP?

DBA does not use a multi-bucket item forecast to drive MRP generation. A multi-bucket forecast typically projects monthly usage for an item over several months into the future. Sometimes blanket POs are used as a type of multi-bucket forecast.

Multi-bucket forecasts are inefficient and lead to over-stocking. It is difficult to distinguish actual short term demand from forecasted demand, which means that jobs and POs are often created in response to forecast amounts even when actual demand is less than forecasted. To be useful, multi-bucket forecasts require rigorous and tedious maintenance, which is often not practical when potentially thousands of items are involved.

Instead, DBA uses a single bucket monthly forecast that does not get confused with actual demand. The item forecast amount is used in conjunction with the item's planning period days to calculate a Reorder Point. A new job or PO only gets generated when actual net demand falls below the item's Reorder Point.

A single bucket forecast is much easier to maintain than a multi-bucket forecast. Not only is it limited to a single amount, it also does not require the same degree of accuracy because there is much less risk of over-stocking or significant shortages. If the monthly forecast is higher than actual demand, MRP will not generate additional supply until the Reorder Point gets exceeded. If the monthly forecast is too low, MRP will generate additional supply and any shortage that occurs will likely be of minimal duration.

For many if not most items the single bucket forecast can be left as is for long periods of time. A good maintenance strategy is to review forecasts on a spot basis whenever shortages occur and to perform a mass review once or twice per year.

← Back