In the previous blog, it was stated that push channel replenishment is driven by two factors:
- First, channel replenishment was calculated by forecasting channel demand at the aggregate level. This task is usually performed weekly or monthly.
- Product is then pushed from the manufacturer or DC to downstream channel echelons using some form of replenishment algorithm based on the forecast.
An example of a push replenishment system appears in the below figure. The supply channel consists of a supplying manufacturing plant, two distribution centers (DCs), and four remote warehouses supplied by the DCs. The movement of historical demand information moves upstream from the warehouses to the plant and replenishment inventories move downstream from the plant to the warehouses. The process begins at the plant where inventory planners determine the total requirements of the channel by aggregating demand and creating the forecasts that they then will use to determine production lot sizes and replenishment order release.
Once production is completed, the replenishment quantities are allocated to the two DCs. In the example, 60 percent of the output of the plant will be “pushed” to DC1 with the remaining 40 percent pushed to DC2. The DCs are each responsible for resupplying two remote warehouses. Based on predetermined criteria, 55 percent of DC1’s inventory is pushed to warehouse 1 and 45 percent pushed to warehouse 2. Likewise, DC2 pushes 65 percent of its inventory to warehouse 3 and 35 percent to warehouse 4. The expectation is that the allocation percents follow closely actual channel demand requirements through time.
The below illustration shows how a push system allocation is used in a complex multi-echelon distribution channel. The figure details the national sales forecast for an item for a large apparel retailer operating in the U.S. Styles are non-repetitive and are available only during the seasonal campaign. Such a period-level forecast is developed by the marketing department using special statistical software that takes past sales history combined with correlational factors and market estimates. The goal of the planners is to allocate the national forecast for the item across a distribution channel consisting of regional and area facilities which, in turn, supply local retail stores.
The top row of the grid displays the national forecast for an item spread over an eight-period planning horizon. The goal is to calculate the quantity of the national forecast to be “pushed” or allocated over the business’s four regions. This is easily done by multiplying each period’s national forecast by each region’s predetermined replenishment percent. Basically, this percent represents the portion of expected demand for each region. For example, the quantity to be “pushed” to Region 1 is calculated by multiplying the national forecast of 15,000 units by 12 percent to arrive at a quantity of 1,800 units in period 1, 1,800 units in period 2, and so on.
The same logic is applied to calculating the inventory requirements at the four channel Areas comprising Region 1. For example, the forecast for Region 1, Period 1 is 1,800 units. The allocation for Region 1, Area 1 is calculated by multiplying the 1,800 units by 40 percent (the predetermined replenishment percentage for Area 1), or 720 units. Finally the same logic is used to calculate the replenishment quantities to be pushed to the four channel Stores. For example, the forecast for Area 1, Period 1 is 720 units. The allocation for Region 1, Area 1, Store 1 is calculated by multiplying the 720 units by 30 percent (the predetermined replenishment percentage for Store 1), or 216 units. Once the national forecast allocation is performed for all channel echelon levels, the inventory quantities are then physically shipped or “pushed” through the distribution channel down to the store level.
In the next blog will be reviewing the pros and cons of the push system technique.