Developing a Long-term Inventory Replenishment Forecast for a Vendor

Vendors will on occasion require a distributor provide them with a long term inventory replensishment forecast, (e.g., monthly requirements for the next six months) which they will theoretically use as a planning device. Hopefully for the distributor, the forecast is non-binding, and the distributor can place the actual orders on a periodic basis as time progresses. There are situations, however, where the forecast is binding. It is then often referred to as a blanket order.

In either case the problem of developing a forecast, or order of this nature, can be a difficult and time consuming job and even then results in a relatively poor reflection of the company’s true needs.

MARS Inventory Replensishment and Forecasting software can provide this requirement very effectively and accurately. The key is to use the “blanket order” feature of the system. (See “blanket order” in the documentation for the exact steps needed to perform this function.)

The net result of this feature is that it provides a series of purchase orders for as many forward periods as you wish. It is actually a simulation of the ordering process as if you were moving through time and placing purchase orders periodically as you would under normal circumstances.

Keep in mind the simulation of the orders will not be exactly the same as if you placed the orders one at a time each separated by the true order cycle. Under actual ordering conditions the inventory replenishment forecasts would change each period as the data was updated with the actual sales. Nevertheless, the blanket order feature will provide the best possible forecast with the data available at that time. The sum of the purchase orders then represents the total forecast for the time span of interest.

One might conjecture that an alternate solution is to do a single order, but with the lead time set at some period that encompasses the forecast period you are trying to predict. (E.g., to provide a forecast for a six month period set the lead time at 180 days.) On the surface this would make sense since MARS would then build an order to cover a full six month period. The problem, however, is that the lead time is part of the safety stock calculation, and therefore MARS will create a very large safety stock because it thinks that it has to protect for an actual lead time of 180 days, which obviously is not the case. The final order is therefore highly overstated.

The blanket order or simulation approach, however, uses the actual lead time and therefore leaves the system after the last order cycle with a safety stock that is exactly what will happen under actual conditions.

To learn more about how Mars can help, see our offer below.