By Dr. Peter Turney and James M. Reeves
Cost Management

Together, the continuous improvement of products and processes and the financial insights that can be gained from activity-based costing (ABC) have become widely recognized as important ways to gain competitive advantage. Continuous-improvement programs are multifaceted, but they usually include steps to eliminate waste, reduce response time, simplify product design, and improve quality. To support these efforts, ABC provides valuable economic information to guide product strategy and process change.

Early examples of ABC demonstrated a superior ability to report more accurate product costs in environments characterized by product variety and batch-size diversity (i.e., in settings where large numbers of different products are manufactured in a wide range of production lot sizes).

The Schrader Bellows case
The Schrader Bellows case, for example, described an ABC system for a plant with over 2,000 products (representing about 20,000 components) and batch sizes that varied from one to tens of thousands. Schrader Bellows committed resources toward many activities supporting the production of batches of product, such as setup, first-piece inspection, and material movement. The traditional standard cost system failed to attribute cost to batches of products. Therefore, the conventional system reported product costs that were markedly different from those reported by the ABC system.

The ABC system provided insight into Schrader Bellow’s strategic capabilities. The company did not have the capability to manufacture many end items in small lots. Like most firms, Schrader Bellows had two basic alternatives in response to this type of insight. One was to drop or re-price unprofitable products and the other were to reduce the cost of producing these products.

Specifically, the company could either change its product strategy or change its strategic capability. Changing product strategy requires either increasing prices or narrowing the scope of products offered. Changing strategic capability, on the other hand, requires continuous process improvement of the value chain or improvement of product design.

Unfortunately, adapting product strategy to competitive pressure is much easier than altering strategic competitive pressure is much easier than altering strategic capability. One of major lessons of the continuous process improvement is that altering product strategy competitive pressure is a short-term solution at best. A firm can give up markets or increase prices. By doing so, the firm may benefit in the short run, but the wellspring of a firm’s strength lies in a superior ability to contribute value to customers. This ability comes from constantly increasing strategic capability.

Impact of Continuous Improvement
Increasing strategic capability through continuous improvement reduces the importance of the factors that made ABC successful at Schrader Bellows. Continuous-improvement programs include actions that reduce or eliminate batch-level activities, reduce batch size, and diminish nonfunctional differences between products.

Continuous process improvement yields reductions in the time required to perform batch-level activities. The length of time required to set up a machine, for example, may be reduced from many hours to just several minutes. This task can be accomplished by doing the following:

• Improved training;
• Eliminating conflicts in workers’ assignments;
• Driving setup time off-line; and
• Placing tools and dies in more convenient locations.

Such efforts can be extremely successful at reducing setup time and costs. Over a five-year period, for example, Toyota reduced the setup time for 800-ton stamping presses from more than one hour too less than 12 minutes.

Some batch-level activities, such as those associated with incoming materials, may be completely eliminated by continuous improvement. The inspection of materials for quality defects, for example, may be eliminated if the vender is able to deliver materials exhibiting statistical control. The following may also eliminate receiving, moving, and storing activities:

• Having venders deliver materials directly to where they are used on the production line; and
• Reducing the distances between adjacent work centers from hundreds of yards to a few feet.

Reductions in batch size. The reduction or elimination of batch-level activities creates opportunities for a reduction in batch size. It is no longer necessary to maintain large batch sizes to spread the high cost of batch-level activities over a large number of units. In the extreme, it is possible to produce and sell in lot sizes of one at the same unit as with lot sizes of one at the same unit cost as with lot sizes of 1,000 or more. This change is a positive development because small-lost sizes lead to improved quality, flexibility, and responsiveness to customers. Process changes can, therefore, have a significant impact on the design of ABC systems, which is a subject discussed in the following pages.

Eliminating nonfunctional design differences. Continuous improvement not only improves processes by eliminating the economic significance of batch size but also improves products. This improvement is accomplished by reducing differences between products that are not the result of functions that customers define as critical. The objective is to design products with minimal nonfunctional differences. Products may be designed, for example, to eliminate nonstandard components and subassemblies that do no add features required by the customer. As a result, imposing a common basic design across product lines can eliminate many individual parts.

As another example, package goods manufacturers produce products in a number of sizes. These varieties of size are customer-driven requirements, so they represent functional product differences. The package design engineer, however, can impose commonality in packaging specifications by using with common rounds. Common rounds refer to the diameter of the container bottom: The key is to accommodate size differences in the height of the container, not the diameter, because package lines must be set up a given diameter. Size changes by diameter necessitate expensive line changeovers whereas size changes by height are relatively simple.

The reductions in nonfunctional product differences simplify the entire production process. Products that are functionally different but virtually identical from a design standpoint share process and support activities. Economies of scale on common activities are realized as production is organized around “families of products” or common subassemblies.

Low-cost product variety. Firms whose continuous-improvement programs are well advanced may increase the variety of their products. The ability to produce in small lot sizes and the simplification of product design reduce the cost of producing a wide range of products. Low-cost product variety (which is essentially the ability to provide customers with virtually customer products) is the ultimate strategic capability derived from continuous improvement.

Having identified the impact of continuous-improvement activity on the firm’s processes and products, how does the ABC system respond to these changes? Does the firm move toward simplified systems in which the cost of identifying activities outweighs any perceivable benefits? These questions can be answered by addressing the impact that continuous improvement has on the following:

• The type of cost driver required;
• The number of cost drivers required; and
• The changes in the cost object.

Type of Cost Driver
One of the fundamental characteristics of ABC systems is the distinction between activities at the unit, batch, and product levels. Figure 1 illustrates the hierarchy. Unit-level activities occur each time that a unit is produced, while batch-level activities occur every time that a transaction affecting more than one unit occur. Product-level activities affect a complete product line and vary in accordance with product code (part number) proliferation.

The cost of batch-level activities must be divided by the batch size to determine unit cost. Similarly, the volume of the product line to determine unit cost must divide the cost of product-level activities. Similarly, the volume of the product line to determine unit cost must divide the cost of product-level activities. For example, the cost of a setup must be divided by the production lot size, and the cost of engineering a new-part drawing must be divided by the part’s volume to determine unit cost. As a result, batch-size diversity and total product volume diversity become important reasons to have an ABC system since the denominators are different across products.

Continuous improvement reduces the importance of batch-level activities and also reduces batch-size diversity (the ratio of batch sizes). These changes eliminate the need for batch-level cost drivers in the ABC system. Cost drivers associated with production batches (e,g., the number of setups or the number of production runs) will ultimately disappear because the cost traced will be insignificant. Likewise, cost drivers associated with orders of components or materials (such as the number of inspections or the number of receipts) should also disappear with the implementation of continuous-improvement strategies.

Product-level activities. Even a firm that has removed batch-level activities will still have product-level activities, which are needed to service an entire product line. If product-level costs exist, ABC systems will be superior to conventional cost accounting systems when the following conditions exist:

1. There is activity diversity across the product line. That is, products consume relatively different amounts of product-level activities. Activity diversity can be thought of as a numerator effect from the point of view of determining product cost.
2. There is total volume diversity across the product line, which is a denominator effect.

The presence of either of these two types of diversity suggests a need for an ABC system to prevent distorted product cost.

Product diversity. Examples of product diversity in the numerator occur if some products require more engineering attention than others do. For example the drawings for recently released parts habitually need greater engineering support than the drawings for mature parts. As a result, the ABC system should trace the engineering effort to the newer parts so that mature parts are not forced to cross-subsidize the recent releases.

In this way, the benefits of the learning curve are related to products according to actual engineering change experience, as illustrated in Example 1.

Example 1. A medical equipment company manufactured and sold two different types of wheelchairs. One wheelchair was sold to individuals who were permanently disabled, and the other was sold to individuals who needed them for temporary use (e.g. for the elderly). The premium for liability insurance represents a product-level cost because the liability exposures differ for these two difference products. The wheelchair for the permanently disabled poses a higher risk because of the more intense and longer-term use of these two different products. The ABC system should, therefore, assign more liability insurance cost to wheelchairs for the permanently disabled to reflect the activity differences.

Vendor-related activities. Some components require more vendor-related activities than others do. Again, activities to establish and prove vendors are usually concentrated on specialized components or new part numbers. These costs should be assigned to the products that require this additional activity. Product-level drivers, such as the number of engineering changes or the estimated effort by vendor, should reflect these differences in the cost of the products.

Denominator effects. Even in the absence of significant batch-size and product diversity, there may still be volume diversity across the products (part numbers); the result will be denominator effects, as illustrated above.

Example 2. Products A and B are members of the same product family. They are quite similar in design and are produced in lot sizes of one. They consume virtually identical types and amounts of activities. (The effort required to maintain the bill of materials and routing sheets, for example, is the same for each product). However, the sales volumes of the two products are quite different, because A is a high-volume and product B is a low-volume product. The cost per unit of B for product-level activities, therefore, exceeds that of A because of these volume differences. This difference will not be reflected in the reported costs of A and B unless a product-level cost driver (such as the number of products) is used.

Unit-level activities: “Superunit”cost drivers.
Advanced manufacturing settings are characterized by a variety of unit-level activities, which are activities performed on the product unit itself. In a printed circuit board plant, for example, these activities include photoimaging, creating multiple layers, drilling holes, and plating. Each of these process activities is used to a different degree by each product, depending on the design of the product.

ABC systems use “superunit’ cost drivers to reflect the different use of these unit-level activities by the products. Examples include the number of holes or number of layers of a printed circuit board. They also include the number of axial insertions, the number of radial insertions, and the number of manual insertions in a circuit board assembly.

Superunit cost drivers replaces traditional unit drivers, such as direct labor hours and machine hours. Direct labor cost often falls to such a low level under continuous improvement that it is no longer even measured in some plants. Machine hours can be used in lieu of direct labor hours but may require developing a system to measure it. In contrast, Superunit cost drivers are usually stored in the manufacturing database and are, therefore, readily available for use by the ABC system at no additional cost.

Superunit cost drivers is common in plants that produce custom products. Superunit cost drivers provide the link between the different characteristics of the products and the multiple processes in the plant. They enable design engineers to model the economics of alternative product and process designs. A printed circuit board, for example, places different demands on process activities, depending on the number of holes or the number of layers.

Number of Cost Drivers
It is reasonable to believe that an ABC system should have fewer cost drivers as a result of implementing a continuous-improvement program. Even in cases where superunit drivers are required, the reduction of batch-level activities eliminates a whole class of cost driver.

Furthermore, reductions in product diversity reduce the need for product-level cost drivers to achieve improved accuracy of product cost.

In cases where the factory is focused on one product or on a small number of similar products, it is common to find product-costing systems that use a single cost driver. These systems rely on the direct tracing of overhead cost to product cells. A single cost driver (such as cycle time) is then used to allocate the cost in each cell to the products manufactured. It has been argued that such a simple system can still report accurate product costs because of the lack of product diversity within the cell.

One concern with a system that has only a few cost drivers is to be certain that it encourages the right behavior and does indeed report accurate product cost. The Portable Instrument Division of Tektronix, for example, implemented a system with two drivers: the number of part numbers and direct labor hours. The purpose of the driver “number of part numbers” was to focus the attention of the design engineers on the cost of part-number proliferation. The engineers responded by increasing the commonality of new products and reducing the division’s part count.

While this initial response was appropriate, it was clear the commonality was only one of several important engineering objectives. Reducing the part count did not necessarily reduce the number of process steps required; not did it necessarily increase the quality of the product. A subsequent version of the system expanded the number of drivers in the cost system to supply the engineers with the more detailed information that they required.

Potential for inaccurate product costs. A system with only a few drivers may also report inaccurate product cots and thus lose credibility. Zytec, a manufacturer of power supplies, used cycle time and supplier lead-time in its new cost system. Cycle time, which is the elapsed time from the arrival of raw material in the plant to the shipment of the product, was used to trace manufacturing overhead to the products. Supplier lead-time, which is the time that elapses from when an order for a component is placed until when it is delivered, was used to trace material overhead to the products.

The objective was to make the cost system consistent with the company’s continuous improvement program. Cycle-time reduction was intended to reduce cost and improve quality and service. Reduction in supplier lead-time was intended to reduce cost and improve flexibility.

The system failed, however, because it did not report accurate product costs. Both drivers measure elapsed time, yet components and products consume activities in ways that are not reflected by time. For example, a vendor with a long lead-time may deliver high-quality components, whereas a vendor with a short lead-time may deliver low-quality components. Similarly, a product with a long cycle time may require little engineering attention, while a product with a short cycle time may require substantial engineering time.

The response to Zytec’s system was quite negative. The cycle time portion of the system confused managers, who were unable to explain the differences in cost from one product to another. Managers also had difficulty understanding the relationship between supplier lead-time and cost. Finally, customers were unwilling to rely on prices based on costs that were reported by the new system.

One-product plants. Even if a plant makes only one product, it may still be necessary to have an ABC system with several cost drivers. For example, an ABC system was introduced in a defense plant that produced “only one product.” However, this “one product” actually consisted of many subassemblies and thousands of components. Subassemblies were sold as spares and for use in customers’ products. In addition, subassemblies were ongoing candidates for outsourcing. In reality, therefore, the plant had many products and required accurate costs for each of these products. A multidriver system was required to provide this accuracy.

Changes in manufacturing strategy. A one-or two-driver system may be vulnerable to changes in manufacturing strategy. One plant, for example, was organized by product cell. The costing system traced costs directly to each cell and used direct labor hours to allocate cost to the products produced in the cell. This design reported accurate product cost as long as only one type of product was produced in each cell.

Recently, however, losses in sales volume had forced the plant to emerge several product cells. The system was too simple to accurately cost cells with multiple products.

Cost Objective
Continuous-improvement programs typically start in a plant with the elimination of batch-level activities and nonfunctional product differences. As the plant becomes more efficient, plant overhead goes down, and opportunities for improvement become smaller. The next step for some firms is to move outside the plant to improve postplant activities, such as marketing and logistics.

Postplant activities. This shift in emphasis outside the plant recognizes the growing importance of postplant activities. In some cases, the cost of these activities equals or exceeds plant costs. Consequently, there are opportunities to target these resources more effectively, to adjust prices to reflect the use of postplant activities, and to continuously improve the performance of these activities.

To take advantage of these opportunities, an ABC system must be reoriented toward a different cost object. The costs of plant activities are traced to the product in ABC systems. Many postplant activities, however, focus on the customer, rather than the product, as a cost object.

Customer-level activities. The postplant costs, therefore, give rise to a new activity level: the customer-level activity. These are activities that do not support units. Batches, or products but that does support customers. In a sense, the cost system is turned on an axis toward this new cost object: The product cost in the plant is the platform to which the customer-level costs are added. The total cost of serving a customer, therefore, is the cost of the products purchased plus the customer-level costs. The result is a cost, system that provides information on the relative profitability by customer or customer segments not just by product.

Customer-level activities are activities unique to serving specific customers or customer segments. Examples include the following:

• Specialized service;
• Logistics support;
• Merchandise support;
• Special receivables terms;
• Inventory buffering;
• Cash incentives;
• Packaging;
• Palletizing; and
• Engineering support.

These support activities act many like product-level activities, except the cost object is the customer rather than a product.

Customer-level cost drivers. Customer-level activities require customer-level cost drivers, which can include the following:

• Number of customers;
• Service level;
• Marketing effort; and
• Specific inventory and receivable requirements.

Naturally, high levels of activity investment in a customer for small volumes result in low profitability. In a sense, the ABC system brings intelligence that has generally been lacking to the continuously improving firm. The question is not what does the customer want from the, but what does the customer value-in other words, what is the customer willing to pay for products and services to make them profitable to both parties.

The extended value chain. A realignment of the cost object toward the customer is entirely consistent with the direction of any continuous-improvement effort: The effort is an extended value-chain perspective. The hallmark of continuous improvement is to provide value to the customer. Increasing value to the customer may require strategic changes as well as continuous improvements in the activities serving the customer.

Customers that demand high postplant activities should not be cross-subsidized by those that do not. If a company maintains its focus only on product cost, it forces an averaging of different customer activity intensities across the product line. Therefore, cost should include both product costs from the plant plus customer-level costs of support provided outside the factory walls. This information can provide insight into marginal opportunities and vulnerabilities in the customer base, as illustrated in Example 3.

Example 3. To illustrate, a small machine shop with numerical control (NC) equipment was operating three shifts per day, seven days per week, but was making very little profit. The major customer-a Fortune 100 companies that the machine shop considered its bread and butter-provided 50 percent of the machine shop’s volume. But the reality was far different, as a careful examination of customer-driven activities revealed. The major customer ordered high-precision machined parts in low lot sizes. The jobs for this customer, therefore, required long setups, intense engineering support, intense NC programming support, intense sales support, high order activity, higher scrapped units, high inspection intensity, and high inventory for this firm. Rather than assigning these incremental costs to the products manufactured for the major customer, however, these costs were spread across the machine shop’s complete product line. As a result, this major customer enjoyed subsidized pricing; the machine shop’s number-one customer was actually its number-one loser.

Strategic and operational changes. The firm’s response to this information was both strategic and operational. The strategic response was to increase price to offset part of the high cost of serving the customer. The operational response was to reduce the cost of serving the customer via continuous improvement. The information to support both of these responses came from the ABC system, as illustrated in Example 4.

Example 4. When a trucking firm looked at services from a customer perspective, its freight rates did not reflect activity differences by customer. The firm developed an ABC system that used the customer as the primary cost object. Examples of customer-level cost drivers used included were pickup-drop off efficiency, backhaul opportunities, number of stops along a single haul, extend of prior notice by the customer, and extent of surge capacity needs by the customer.

The trucking firm uses the information from its ABC system to change the way it prices its services. The traditional approach to pricing was to set a standard rate to haul freight from one location to another. Under this approach, the rate from city X to city Y was the same regardless of activity intensity.

The new approach to pricing is based on customer-service requirements as defined by the ABC system. The new rates ensure that customers that receive low-cost service pay only for a share of the activities they use. Customers that are expensive to serve share the cost of the additional activities used.

Conclusion
ABC systems in traditional batch-oriented manufacturing environments are likely to report large shifts in the cost of the products. This additional accuracy is a major source of value of these systems because it reduces the cost of strategic errors. For example, more accurate product costs reduce the likelihood that prices will be set incorrectly or the unprofitable products will be sold aggressively.

However, continuous-improvement programs reduce the scope for improving the accuracy of product cost because overhead associated with batch-level activities goes down, batch size diversity is reduced or eliminated, and product diversity may initially be reduced. These changes affect the design of ABC systems in terms of the type of cost drivers used; the number of cost drivers used, and evens the cost object.

Continuous improvement changes the type of cost driver found in ABC systems. Batch-level drivers are not used because of the absence of batch-level activities and batch size diversity. Product-level drivers are prominent because of either continued product diversity or sales volume diversity. Superunit drivers are found in advanced manufacturing settings where custom products are produced on multiple processes.

Continuous improvement also changes the number of cost drivers found in ABC systems. The number of cost drivers may go down because an entire class of cost driver-the batch-level driver-is eliminated. It also may go down because, in some cases, product diversity is reduced. It is common, therefore, to find product costing systems, in continuous-improvement situations that use as few as one or two cost drivers.

It may still be necessary, however, to use multiple cost drivers: Too few cost drivers may focus attention on an incomplete set of improvement requirements and, therefore, interfere with the improvement process. Too few cost drivers may also result in inaccurate product costs and may be vulnerable to changes in manufacturing strategy.

As continuous improvement proceeds, the focus of improvement extends down the value chain from the product in the plant to the customer. This extension recognizes the significance of postplant, customer-level activities and their associated cost. The ABC system, therefore, uses a new cost of customer-level activities consumed to the cost of products.

Under a continuous-improvement program, the need for ABC systems to support marginal evaluation of customers and customer segments becomes increasingly important. Customer cost provides new insights into both customer segments become increasingly important. Customer cost provides new insights into both customer and segment strategy and leads to improvements in the performance of customer-level activities.

About Cost Technology

Cost Technology designs and implements solutions to support fact-based decision making.  These solutions transform existing data into strategically meaningful and actionable insights in areas such as cost and profit management, financial forecasting, analytics, and performance management.  Since 1991, the firm has helped public and private sector organizations across the world improve performance by creating the knowledge needed to make every decision count.