Tradeoff Curves

One of the fundamental tradeoffs in supply chain management is that between inventory levels and customer service. For any given supply chain, increasing the level of service (product/spare part availability) typically means higher levels of inventory. Most companies have discovered their "best place" on the curve, depending on what their customers require and what their competition offers. However, supply chain strategies can shift the entire curve, lowering your inventory levels without adversely affecting your customers (or the reverse, improving customer service levels with no increase in inventory). How might this work? Through effective supply chain management you may be able to reduce lead times. This would shift the curve to the right, speeding up customer response times without raising inventories. Module SCM106 reviews a strategy called postponement, or risk pooling, that can lower the curve, allowing you to maintain (or enhance) service levels with less finished-goods inventory.

This tradeoff curve provides a perfect example of how silo behavior (in which functional areas lose sight of cross-functional optimizations) can cause problems in supply chains. One of the first steps in improving a supply chain is making sure that organizational responsibility for inventory levels and customer service are appropriately managed. These two responsibilities should not be separated - in fact, they should report to the same desk. Doing so enables a company to set expectations and properly manage this tradeoff, without costly swings from one place on the curve to another as different functional groups "fight" for either lower inventories or higher service.

Postponement

Product design often plays an important role in supply chain management. Many products come in different varieties - meeting global demand for variety by holding multiple Stock Keeping Units (SKUs) of similar products can require vast inventories. By redesigning the product so that more inventory can be held in a customizable, "lowest common denominator" form, inventory can be reduced.

This powerful supply chain management strategy is known as "postponement", or "risk pooling"; postponement of the point of product differentiation, or pooling of the risk of specific SKU forecast error. On the left side of the figure below, the product is built for a specific country market before it is even shipped out of the factory (a costly decision if the unit does not sell, and particularly so if there are shortages of a similar product in other countries). On the right, the product has been redesigned so that localization can occur as close as possible to the local market.

Module SCM106 provides an in-depth example of postponement, which lowers the inventory/service tradeoff curve (covered in module SCM101), allowing you to maintain (or enhance) service levels with less finished-goods inventory.

A recent study¹ concluded that companies using postponement strategies in their supply chains typically showed significant improvements in the following performance measures:

• More accurate demand forecasting
• Reduced inventory costs
• Higher customer satisfaction
• Increased revenue

The reason sales (demand) forecasts become more accurate is that postponement changes the entity to be forecasted, typically from item-level (SKU-level) sales to aggregated quantities such as sales by product family or group. Thus, while forecast error at the SKU level may actually be unchanged, there is a distinct benefit to postponement, since it is virtually always true that forecasting aggregated quantities (e.g., sales of a product family) is easier than forecasting detailed quantities (e.g. SKU-level sales). With postponement, the new product design allows managers to make decisions based on the more-accurate aggregate forecast rather than less-accurate SKU-level forecasts.

Inventory costs are lower because there are fewer overstocked products, customer satisfaction is higher because there are fewer stockouts, and revenues are higher because there are fewer lost sales.

However, the survey also showed that many firms have not yet implemented postponement along with their supply chain management initiatives. The major stumbling blocks are:

• Lack of knowledge about benefits of postponement and costs
• Belief that their product or technology limits postponement
• Organizational misalignment (silo thinking)

The first two stumbling blocks can be solved through a greater understanding of postponement at all relevant levels of the company, including finance, manufacturing, sales/marketing, and engineering/design. Regarding organizational misalignment, we have noted elsewhere in our modules that as management discovers the real benefits to postponement, the organizational structure often becomes more accommodating to ensure success (this is true for many cross-functional supply chain initiatives). In the best cases, metrics, or performance measures, on which various departments are evaluated would shift to reflect a more cooperative environment; one in which corporate (and entire supply chain) performance is paramount.

¹ "The Adaptive Supply Chain: Postponement for Profitability" by Oracle Corporation (Nasdaq: ORCL), Cap Gemini Ernst & Young U.S. LLC and APICS; see http://www.oracle.com/corporate/press/2436365.html.

Chain-Wide Performance Measures

Typically, performance measures for supply chains are applied at the company level; one company's inventory turns might be compared to a competitor's inventory turns. However, a new concept is to apply such metrics across the entire chain to see how well the whole chain is functioning. Why are chain-wide metrics important? If you look over your entire supply chain, you get a much better view of how well you are performing relative to your competition in the eyes of the end-customer. Your partners in your supply chain might have a significant adverse impact on the delivery of goods, no matter how well optimized your own inventory/service levels are.

Our total supply chain may be drastically inferior to that of a competing supply chain, but we may be unable to observe that if we only compared our local node in the chain (e.g. our factory) with that of our competitor. Let's re-visit our inventory/service tradeoff curve in the figure below; the left side of the figure shows that our inventory/service tradeoff curve is very similar to our competitor's. The right side of the figure, measuring our entire supply chains, reveals that our competitor's overall chain achieves better response time with lower inventories. Given that we know that our company/node is performing well, what is then implied about the performance of our partners in our supply chain? What actions might that suggest?

For example, consider two motherboard manufacturers that each sell to a different PC maker. Both board manufacturers might have excellent, even similar, inventory/service levels, but if one of the PC makers is consistently late shipping product, customer satisfaction will drop - as will demand for their suppliers' motherboards. The motherboard manufacturer has a vested interest in understanding the weak points in the overall supply chain, and may consider either working with the PC manufacturer to solve the delays or selling to a different PC manufacturer. This is often described as "competition across supply chains" rather than individual companies. Internet technologies are providing for real-time collaboration in forecasting, sales information, and product design to make this a reality.