The Real-Time Supply Chain
Demonstration model 3: Postponement

To cope with demand variability and mass customization, companies started to implement make-to-order strategies. But with increasing competition and customer demands, delivery lead times have become increasingly important. In this light, a new concept called postponement arose.

Postponement comes in two flavors: process and product. Process postponement is a movement of the push pull boundary as close to the finished end product as possible in the manufacturing process. A generic component is produced in the initial stages of product, which is then customized in a later stage. With product postponement, the customization step is moved as late into the production process as possible.

Furthermore, this customization step doesn't necessarily have to be done at the manufacturer level. And as can be read in the Xilinx article, this can be even after a customer bought the product! Another example is Hewlett-Packard. HP manufactures printers for worldwide distribution at its Vancouver base. Due to demand variability, it was faced with frequent inventory imbalances (e.g. too many "Spanish" printers while the German version was out of stock). After a study, Hewlett-Packard redesigned the printers in such a way that the country-specific components such as power supply and owner's manual could be added in the distribution centers. This concept is known as demand localization [From: Handfield, R.B., Nichols jr., E.L., Introduction to Supply Chain Management, Prentice Hall, 1999].

Postponement has the advantage of responding to the uncertainty of demand in particular markets through purposeful delay and thereby avoiding the risks of inventory accumulation and obsolescence. Another possible advantage is that production is more efficient, as part of the production capacity can be shared among products.

Other benefits are the reduction in supply chain lead times, inventory levels and system costs, while simultaneously making it easier to manage system resources. If the manufacturing steps are modified to delay the final differentiating of the product until the final stage, semi manufactured products can be produced based on the long-term forecast while the final differentiating is driven by actual orders.

One of the most famous examples of product postponement is Benetton:

Benetton is a major supplier of knitwear. The nature of the fashion industry is that consumer preferences change rapidly. Because of long manufacturing lead time the store owners had to place order up to seven months in advance. This left little room for flexibility to respond to the changing market. To address this issue Benetton revised the manufacturing process by dyeing the garments after the sweater was assembled instead of dyeing the wool. (Simchi et al, 1999)

To show the effect of a supply chain with and without postponement, the model contains two very similar supply chains: one with product postponement and one bases on conventional make-to-stock. Each consist of 1 customer, 3 retailers, 1 manufacturer and 3 suppliers.

Figure 1. Postponement supply chain

The end product, a sweater, is available in three different colors: red, green and blue. A sweater is produced using wool, buttons, zippers and coloring. The manufacturer that uses postponement will make with wool, buttons and zippers semi manufactured products: sweaters without coloring. After an order the sweaters will be dyed in the desired color. The traditional manufacturer directly dyes the desired color and keeps inventory for each of the three colors.

Model: demo3.jar. Click here to start the Java-based model from your browser

Tips:

• Use the "play" button to start the simulation
• There are two animations, one of the "normal" supply chain, and one of the "postponement" supply chain.
  You can get access via the menu of the main window.
• Press F2 to get access to the graphs, select one of the graphs in the left window by clicking on the "+", and drag the graph into one of the cells of the viewer. Graphs also come in two pairs: normal and postponement.

Articles:

Brown, A.O., Lee, H.L., Petrakian, R., Xilinx improves its Semiconductor Supply Chain using product and process postponement, Interfaces, July-August 2000, pp 65-80.

Questions:

1. What are factors that influence whether a supply chain will move to make to order or make to stock?
2. Compare the two supply chains for a number of key performance indicators over a run of around 6 months.