06.3 Configure And Manage Supply Chains To More Effectively Address The Challenges Associated With Inventory Management (Part 3 of 3) (DDPSC06.3)
06f. Who defines the required service level and how should it be defined?

The customer defines what the required service level is, and they usually do so through their buying actions - complaining when product is not available, buying more than what they need when product is available, returning to buy the product the next time they need it, or, when the product is not available going elsewhere to buy what they need, and sometimes not ever returning. If service is really important and your service level is very high compared to the competition, they may even recommend you to others. In other words, service levels should be defined in accordance with the impact they have on the bottom line, which means fill rates that are less than 100 percent and don't result in lost sales may very well be acceptable fill rates.

Inventory buffers established using Time Based Demand Patterns are sized in accordance with the supply and demand variability they are intended to contain, in conjunction with the corresponding investments in inventory and risk of being out-of-stock. From the supplier's perspective, it is the tradeoff between the investment in inventory and the risk of being out-of-stock that defines what they view as an acceptable fill rate.

Using Time Based Demand Patterns enables suppliers to better quantify the outcome in advance. See DDPSC06.1 Figure 5. An example of a Time-Based Demand Pattern, below.


06g. What kind of policies, measures and information exchange issues can cause replenishment based inventory buffers to fail to perform?

As we discussed in DDPSC02, Identifying and Addressing Organizational Measurements, and DDPSC04, Policies, Measures, and Organizational Alignment Issues, policies and measures that view the performance of a link within the supply chain as if it were an independent link (as opposed to an interdependent link configured to deliver a certain level of overall supply chain performance) can reduce the effectiveness of replenishment-based inventory buffers. It is this same "independent versus interdependent link" perspective that often underlies the non-collaborative information exchange position of some organizations, as well as the ineffectiveness of some collaborative information exchange processes.

On the supply side this shows up in the form of resource efficiency and utilization, volume, length or weight incentives, overtime that has become part of the standard of living. On the demand side this looks more like generating sales by selling the product "on-sale," through promotions or volume discounts, all of which result in delays (timing) and batching (spikes) of both product and information flow. Any and all of these types of policies, measures and information exchange issues can reduce the effectiveness of the TRR in containing supply-side variability and in the effectiveness of the corresponding inventory buffer in containing demand-side variability.

For example, when we talk about replenishing inventory according to actual usage, the link responsible for inventory management of the downstream buffer might well be required to create, process and send order information daily for the quantity of inventory used or sold each day back to the upstream feeding buffer or to the upstream manufacturing source for subsequent replenishment within the TRR. From the perspective of "efficient use of resources," the daily processing of replenishment orders at the downstream buffer may not actually be viewed as being efficient, even though from the overall system perspective it is. As result, the resources at the downstream buffer may decide that it is more "efficient" or more "productive" for them to accumulate replenishment orders and process them in larger batches (e.g., weekly). The effect of batching of replenishment orders is a time delay in information exchange upstream combined with a larger order quantity. The time delay consumes time within the TRR, making it less effective in containing supply-side variability, which in turn makes the size of the inventory buffer less effective in containing demand-side variability, which reduces the performance of replenishment-based inventory buffers. The net result is that the TRR's need to be increased to accommodate for the time delays associated with the order batching, the size of the downstream inventory buffers will likely need to be increased and the demand data transferred upstream no longer reflects actual POS variability.

The other question that often comes up has to do with the assumed increased cost of creating and processing more replenishment orders. While the daily processing of replenishment orders may increase the amount of data transferred, the assumed cost of actually processing more orders seldom materializes, since the more often a person performs a specific task the more efficient and productive they actually become and the time required per unit tends to go down. In the end, the speed of work and information flow increases while the productivity of the resources involved goes up. The bottom line is cost and investment in inventory to operate goes down, while service levels and sales go up.

Next: Configure and manage the supply chain to more effectively address the challenges associated with achieving Operational Performance (DDPSC07)

Comments or questions? Email us at info@goldratt.com.