A misconception about regions is that it is a summary view of all the branch (location) information that forms part of that region. Commonly asked questions include:
Why is the Excess at my region not the sum of all the Excess stock at my individual locations in that region?
Why is the order quantity at the region not just the sum of the order quantities at the branches?
Why is my risk at the region lower than the average of risks at the individual locations?
The reality is that the region only consolidates the sales history, stock on hand and forecasts (if “consolidate forecasts” was ticked when creating the region) across all locations. From there, the region calculates its own required safety stock from its own set of risks, which will be different to simply averaging the safety stock and risk across the branches. It assumes the region is one location with one safety stock value and one set of risks, whereas the branch model assumes a different safety stock and risk in every location.
Picture this!
Assume the following scenario:
Branch 1 has excess stock of 1000 units while branches 2 to 4 have stock-outs of 200 units each. The region, however, is not recommending an order to be placed. Why? Because the excess stock at branch 1 means that at an aggregated level, no additional stock is required. This means that branches 2, 3 and 4 will remain in a stock-out position.
A region assumes automatic re-distribution of imbalances in the supply chain.
When/where is ordering in a region useful?
As seen in our example above, a region considers the total “stock in chain” and offsets these imbalances in the supply chain. A region will not order more stock if there is sufficient stock for the region. The DC model does not account for “stock in chain” and will order to a distribution center (DC) every time there is a requirement at a store. Using a region is a better financial strategy than using a DC model as excess is re-distributed to where it is required, resulting in a lower stock holding compared to the DC model.
The region model is the fastest way to reduce inventory in a company where high stock holding is their biggest problem.
A region is useful when one can assume, from a planning perspective, that all base locations, making up the region, are physically in one place. In other words, where transfer lead times between the base locations are typically negligible.
The sales history in the individual locations (branches) may be volatile, resulting in higher risk. Because the region used the consolidated sales history across all the locations, the sales history at the region is a lot less erratic. Smoother sales history results in more predictable, more accurate forecasts. As a result, the risk calculated in the regions will be far lower than that of the individual locations. This results in a lower level of safety stock being calculated at the region and thus a lower level of safety stock being assigned to the individual locations. Short answer - A region reduces the amount of capital tied up in safety stock by reducing the risk.
Having a region allows you to:
Manage one set of risks, instead of multiple risks at the various locations.
Reduces the admin time associated with managing the risks across the individual locations.
Seasonality becomes more apparent at a regional level when compared to individual locations.
Using the disaggregation feature allows you to aggregate the sales history across the individual locations, calculate a fitting forecast based on the smoothed sales history and disaggregate it down to the individual locations according to their percentage contribution to the total sales history at the region.
Using disaggregation means that any global forecast adjustments will also be split accordingly across the various locations. This means that a 25% spike due to a promotion will reflect more realistically when applied to the region and then disaggregated when compared to the individual forecasts.
When should a region not be used?
For the region model to be beneficial to a company there needs to be reactive and cost effective internal re-distributions to be able to move those imbalances to where they should be. If that’s not in place, it could have the reverse effect and harm the company by not ordering to locations that require stock when another location has excess.
A Distribution Center (DC) should not form part of a region. If it does, you’ll lose visibility of goods in transit.
Do not make use of regions if the supply chain includes items with multi-level bill of materials (BOMs.)
Always consider alternatives before creating a region. Consider what the real problem is that needs to be solved. Could the problem be solved by aggregating data or locations in the data import instead? Having regions may extend the data import and post time, so configure them with caution and only when really needed.
Final considerations
Because a region is an idealistic view which assumes automatic re-distribution, it is important to consider the following when deciding whether or not to use a region:
Excess at a region will be understated
Achieved fill rate percentage will be overstated
Recommended orders may be lower
However, making use of a region may be a practical starting point in some cases.
You might also be interested in searching the following: