Supplier Selection Scenarios

When designing a new product, several levels of supplier selection decisions must be made. First, the organization must decide whether a part or subsystem (we will refer to either as a “component”) is designed uniquely for the product or selected from a set of standard components. (In some cases, the standard component may have been previously designed uniquely for a different product.) Once the component has been specified, the organization must decide who supplies the component.

There are five main supplier selection scenarios shown below. This calculator will consider decisions under the fifth scenario.

1. If the component was uniquely designed, the company may decide to manufacture internally
2. The company may also decide to outsource manufacturing
3. If the component was selected from a “sole-sourced” supplier, there are few options — the company must try to negotiate the best terms possible
4. If the component is a pure commodity, then company may select from one or more vendors based on various supply chain performance criteria
5. If similar versions of the component can be obtained from multiple vendors, the company must trade-off technical and supply chain performance criteria

The first four scenarios are predominantly driven by procurement and supply chain functions with only minor input from engineering. The part specifications have been given and the design is complete.

The fifth scenario, however, requires close interaction between engineering, procurement, and supply chain organizations. The Supplier Selection Calculator was designed for this fifth scenario.

Trading-off Materials Costs, Technical Performance and SC Performance

Engineering organizations are usually motivated to maximize technical performance and minimize material costs. This calculator will help you trade-off increased material costs against lower inventory costs. Decreased technical performance (if relevant to the decision) may or may not be a “show-stopper.” There are two situations when engineering might be willing to accept higher material costs and lower technical performance.

First, if the technical performance of the component relates only weakly to overall product performance. Usually, these types of components must only carry-out simple functions at or above a minimum level of performance. Better performance does not increase overall product performance and is simply “wasted.” You may be able to substitute a lower performance component so long as it still meets the minimum thresholds.

Second, if the technical performance of the component can be compensated for by “tuning up” another component that has “excess performance.” In most systems of moderate complexity, there are sufficient interactions between components that this type of compensation may be possible at little or no cost.

For these same two reasons, you may be able to use common parts and subsystems across multiple products. As shown in the Commonality Calculator, this is also a way to reduce inventory costs.

Inventory Impacts of Supplier Performance ?

As mentioned in the Inventory vs. Material Costs Calculator, there are four primary supplier factors that directly impact inventory.

First, lead time. The longer the lead-time, the farther out the forecast period, and the greater the chance that the amount required will be different from the amount that actually arrives. In addition, the shorter the lead time, the shorter the pipeline inventory. With long lead times (e.g. ocean freight from Asia to North America), pipeline inventory can be significant. Generally, if you pay for the components when you place the order, you are responsible for pipeline inventory. If you pay for the components when you receive the order, you are not. Of course, payment terms and return policies complicate this distinction.

Second, lead time reliability. The more uncertain the lead time, the higher the inventory buffers that are required. You don’t need to hold as much inventory when suppliers perform reliably.

Third, delivery frequency. The more frequent the deliveries, the shorter that components sit around before they are built into products. If deliveries are every month, then the average working inventory is about 2 weeks (about 4 weeks at the beginning of the month and 0 at the end, so an average of about 2 weeks). If deliveries are every week, then the average working inventory is ½ week. Frequent deliveries are usually obtained from local suppliers or those suppliers who establish programs like just-in-time and vendor hubs.

Fourth, component yield. In some cases, arriving components may be defective or otherwise unusable in a product (e.g. the component has wider tolerances than the design or manufacturing process requires). Yield has three main effects.

1. It increases effective lead time. Only a fraction of the order will take the normal lead time. The defective components must be reordered, and then some of those replacements will be defective and need to be reordered, etc.
2. It increases material costs (assuming the supplier doesn’t send free replacements.
3. It increases ordering and rework costs for ordering replacements and fixing the assemblies (if the defects are not caught in an incoming test).

See below for a more detailed description of all the parameters used in this calculator.

To use the calculator, simply change one or more of the blue numbers and then click the Calculate button.

To Introduction Reload Page

Note: These calculators were created to facilitate rapid analysis of DFSC decisions. Although these calculators make simplifying assumptions, they have proved useful in practice. For complex trade-offs of additional cost factors or for especially important decisions, you may want to perform a more detailed analysis.

Inventory Parameters

• Demand: Mean demand or consumption at an inventory stockpile (units/week). In this calculator, we assume demand of all items is the same.
• Forecast error (fe): Standard deviation of the difference between forecasted and actual customer orders, divided by the mean demand. This ratio is sometimes also called coefficient of variation (CoV). In this calculator, we assume forecast error for all items is the same. Lead time (L): Mean lead time to replenish product into an inventory stockpile (weeks). In this calculator, we assume that lead time for all items is the same.
• Lead time uncertainty (lu): Standard deviation of replenishment lead time divided by the mean lead time. In this calculator, we assume no lead time uncertainty.
• Target service level (SL): The desired percentage of demand periods where there are no stock-outs. It is sometimes called availability rate. Depending on the target service level, safety stock is scaled up or down by a factor “k.” In this calculator, we assume that target service level for all items is the same.
• Review period (R): The period of time in-between physical review of inventory stockpiles (weeks). It is assumed that replenishment orders are placed after each inventory review. For continuous review systems, R is zero. In this calculator, we assume review period for all items is the same.
• Inventory cost (c): The current cost of the inventory, including material costs and any incremental value-add ($/unit).
• Annual inventory holding cost (h): Expressed as a percentage of inventory cost, the annual inventory holding cost typically includes financing, storage, devaluation, and scrap.
• Inventory weeks of supply: Inventory units divided by the mean weekly demand (WOS).

Sixty to eighty percent of all costs are locked-in during the design phase.

Are you considering the supply chain impacts of your product designs, or do inventory, manufacturing, and transportation costs come back to bite you? Most design for supply chain (DFSC) decisions slow to a crawl as companies struggle to quantify the costs and benefits of the various alternatives. There’s a better way. Leading organizations use “rough-cut” techniques rapidly trade-off important cost and benefits like material, inventory, manufacturing, distribution, and lost sales costs. The reason? Decisions come up constantly, and there isn’t always time to collect the data you need to complete a detailed analysis.

The principles of effective design for supply chain are to:

• Utilize cross-functional teams
• Capture intangibles through a qualitative assessment
• Quantify what you can using rapid analysis techniques
• Perform sensitivity analyses on your assumptions

These web pages provide you with several very effective rapid analysis techniques.

The Role of Inventory Calculators

One of the most common trade-offs that design teams make is:

“Should I pay $X more per unit in material costs to enable future inventory reductions through
(commonality, postponement, shorter supplier lead times, etc.)?”

Design teams find this question difficult to answer because they don’t always have experience in quantifying the drivers of inventory. No problem. These web-calculators capture the drivers.

The calculators:

• Don’t require a lot of data
• Help you quantify the inventory benefits of things like increased part commonality or decreased supplier lead times
• Give you answers in $/unit so you can make direct trade-offs against material and assembly costs

Each of the three calculator pages are described below. Click on the button that best describes the decision you’re facing.

When you are making trade-offs that impact inventory, you first need to understand the relative importance of inventory costs and material costs. As a way of estimating the size of the opportunity, you might ask questions about the extremes.

For example:

• If you could eliminate all inventory, how much more would you be willing to pay in material costs?
• If you could save $X/unit in material costs, how much more inventory could you afford to stock?

One approach for reducing inventory levels while maintaining end-product variety is to use common parts and subsystems.

Another approach for reducing inventory levels is to select suppliers that have short and reliable lead times, frequent deliveries, and low defect rates.

Commonality reduces inventory, because the forecast error for aggregate demand is lower than that of individual-item demand.

With high performance suppliers, you don’t need to carry as much inventory in-between deliveries or to buffer their uncertainty.