Accounting for Continuous Improvement
IN 1981 THE PORTABLES DIVISION of Tektronix faced a problem that is familiar to many U.S. companies; the Japanese had entered Portables’ market for electronic meaurement instruments, creating intense competitive pressures. The Japanese priced their products substantially below the prevailing market. They were able to capture market share despite limited quality and performance.
The Japanese appearance was seen as a threat to the business of the entire company. The immediate competition was limited to Portables’ low-priced products, but it seemed likely that competition would spread to Portables’ higher-priced products and eventually to the products of other Tektronix divisions. After all, the Japanese had entered other U.S. markets in a similar fashion and had succeeded in severely diminishing the role or presence of competing U.S. companies. Clearly the beachhead had to be contained, and Portables was the front line.
Portables’ charge was to “stop the Japanese, and it’s okay if you lose money doing so.” They did so by matching prices and thus denying the Japanese a price advantage. Portables was able to slow down the loss of market share, but only at the cost of heavy financial losses.
It was questionable whether this strategy would work over the long term. Portables’ costs were clearly higher than the prices they were now charging for their instruments, and probably higher than the costs of the Japanese who had established the new price level. The rest of the company was profitable and was willing to carry Portables—but for a limited time only. The Japanese were improving the quality of their products and moving up to the middle of Portables’ product range. Clearly something had to be done—and quickly.
In 1983 Portables adopted a program of continuous improvement designed to blunt the Japanese competitive threat. The results were immediate and dramatic. By 1987, cycle time had dropped from an average of twenty-five weeks to seven days. Inventory levels had dropped by 80 percent, while sales had increased. The number of instruments in work-in-process had dropped from 1,500 to 125. Floor space occupied by the division had dropped by more than 50 percent. Five products that had previously been built on separate lines were now built on one line. The number of vendors had dropped from 1,500 to fewer than 200. Quality was up, and more than 70 percent of sales were delivered within two days of the customer order. Above all, market share was maintained, and profitability was excellent.
Portables’ strategy was to set the standard for performance and price that the rest of the industry tried to emulate, and continuous improvement was a requirement of that strategy. Competitors analyzed each generation of Portables’ products and then tried to produce it less expensively, lb survive, Portables had to keep moving their products toward more performance for the dollar. Such a strategy required continuous attention to manufacturing, engineering, and marketing performance.
Innovations in accounting systems and approaches were a prerequisite to Portables’ successful turnaround. This article tells the story of how one division successfully changed its accounting and thus ensured the success of a dramatic change in its manufacturing strategy.
What Is Continuous Improvement?
Before we look more closely at Portables, though, we should define continuous improvement as the relentless pursuit of improvement in the delivery of value to the customer. It is a philosophy that states that production, engineering, and marketing excellence require ongoing attention and learning.
Manufacturing excellence involves a concerted effort to reduce inventories, improve quality, finish the product at the time required by the customer, and reduce cost through eliminating scrap and other wasteful expenditures. Just-in-time (JIT) inventory flow eliminates or reduces buffer inventories and links suppliers, parts inventory, and production to final demand by “pulling” parts and work through the plant according to customer orders. It also involves shrinking manufacturing cycle time as a means of improving quality and responsiveness to customer demand and reducing cost and inventory. Total quality control focuses on “making the product right the first time” and helps eliminate costs associated with poor quality such as scrap, rework, and inspection.
Engineering excellence requires simplified and common product designs that facilitate continuous improvement in manufacturing. Simplified designs, such as those that incorporate modular designs with reduced part counts, make it easier to reduce inventory and cycle time. Common product designs permit a mix of models to be produced on a single production line—a flexibility that reduces customer lead time and improves production stability.
Marketing excellence requires understanding the needs of the customer and being able to meet those needs in a cost-effective manner. Marketing’s job is to communicate the needs of the customer to production and engineering and to ensure that reduced cost as well as improved quality, delivery, flexibility, and design provide competitive benefits in the marketplace.
The Corporate Worldview
To understand how to bring accounting systems up to date, it is necessary to understand why they are poorly adapted to the needs of today’s competitive environment. After all, these systems performed well for many years—why are they broken today?
Today’s accounting systems are like the Cadillacs of the late 1950s. These cars sold extremely well. They were symbols of success, and they were a standard for excellence throughout U.S. industry. They clearly met the needs of automobile customers of that era. Thirty years later, however, there has been a major change in the competitive environment. This change has given us a new “world-view” of how to design, produce, and market an automobile. Today the 1950s’ Cadillacs seem garish and excessive.
Understanding this shift in perspective is the key to understanding the obsolescence of accounting systems. The worldview is a set of beliefs about how to compete successfully given a certain set of competitive conditions (see Figure 1). These beliefs include how the product should be designed, how it should be produced, how it should be sold, and what role accounting should play.
The purpose of accounting is to develop systems that reinforce the activities necessary for the success of the business. It will do this well if these systems are consistent with the worldview, assuming that the worldview is appropriate.
The objective of Portables’ accounting department in the 1970s (and the design of its systems) was consistent with the worldview of that time (see Figure 2). The department’s primary objective was the preparation of external financial reports. This task was accomplished by focusing on inventory valuation, control over transactions, and asset integrity. It was an appropriate objective in a world where monthly net income was the yardstick of success.1
Portables’ accounting systems in the 1970s were generally consistent with the manufacturing philosophy of the time. The use of work orders, for example, was consistent with batch production. Labor efficiency standards reflected the focus on the worker as a source of productivity improvement and the belief that increasing volume was the key to improving efficiency and increasing sales. An inventory reporting system kept track of the buffer inventories used to maintain high labor utilization and to improve throughput.
Unexpected circumstances in the late 1970s caused this worldview to change (see Figure 3). At Portables, the impetus for change was the shock of Japanese competition. The firm made dramatic changes in the way it manufactured, designed, and sold its products. A new worldview emerged: an emphasis on continuous improvement of the ability to serve customers’ needs.
Continuous improvement required a new philosophy for manufacturing, engineering, and marketing. Manufacturing philosophy emphasized new approaches such as people involvement (pushing decision making down the organization and using teams to solve problems) and visual control (visual identification of problems in the production process and their real-time resolution). Engineering philosophy emphasized design for cost, quality, flexibility, and overall customer value rather than for mere improvement of the product’s engineering features. Marketing philosophy focused on matching the needs of the customer with Portables’ ability to deliver value, instead of on expanding the number of product variations as a means of increasing market share.
This new set of philosophies required eliminating obsolete systems, changing systems that would impede continuous improvement, and introducing systems that would encourage improvements. Table 1 summarizes the demands on the new accounting systems.
Creating Change
Accounting innovation requires radically changing the role accountants play and the information they provide to management. This degree of change is difficult to achieve, and there are few published accounts of innovation in accounting in the new manufacturing environment. It is therefore significant that Portables was able to innovate successfully, enlisting management’s full support and contributing to the turnaround of the business.
A New Managerial Team.
The crisis of 1983 prompted Tektronix to create an environment at Portables in which change could take place. Tektronix did not believe the existing management team was capable of dealing with the crisis. A new management team was brought in to turn the business around. This team was committed to continuous improvement and immediately instigated major changes in all functional areas of the business.
The new team had no vested interest in existing management approaches and was intolerant of systems that had no value to management. Its independence was encouraged by a mandate for change from Tektronix’s top management, as well as by the move of Portables’ plant to a point thirty miles away from company headquarters.
A New Accounting Philosophy.
One aspect of this sweeping change was that accounting joined production, engineering, and marketing as a key member of the management team. This role change was demanded by the new managers, who considered it a means of ensuring a supply of information that would support the drive toward continuous improvement. Accounting went from being a watchdog to being a change facilitator.
This focus on information for continuous improvement encouraged accounting to redesign its systems and to develop new financial and nonfinancial measures of performance. A dual reporting system allowed “management costs” to coexist with the “financial costs” that followed corporate procedures and were used for inventory valuation.
Building Trust.
When the new team came in there was no hard evidence that a new approach would succeed, so it was necessary to communicate a sense of direction and vision to the employees. The key to doing so was establishing a high level of trust. Before 1983, many employees had been fearful, suspicious, and resentful. Between 1983 and 1987, trust was established via a participatory management style and a team approach to problem solving. The new managers shared a great deal of information with employees; only payroll and sensitive strategic data were kept secret. Training received special attention. Fortunately, the early changes were successful, and this helped to build trust.
Managing Excess Capacity.
Successful implementation of continuous improvement invariably produces excess capacity. Managers must identify and marshal excess resources if they are to save the costs associated with that capacity; profits will not improve unless these costs are removed or absorbed elsewhere. If reduced costs are not visible to management, then resources for alternative work will not be made available.
But this removal or absorption must be carefully managed so that trust, teamwork, and drive are not damaged. After all, you can’t expect people— accountants or nonaccountants —to eliminate the activity that is keeping them busy if they will be unemployed as a result.
Portables was initially able to transfer the extra capacity to other parts of the company. Extra workers were placed in other positions—at the time, other parts of Tektronix were growing and could absorb these workers. But when the company could no longer absorb Portables’ excess capacity, it was necessary to create new opportunities. Production that had previously been sourced outside the division was brought back inside, and new products were added to the line.
New Faces.
Accounting innovation requires leadership from individuals who are creative, willing to experiment, and able to work with manufacturing. Portables relied mostly on new staff for this purpose—typically on MBAs fresh out of school. Potential staff members were interviewed by people in manufacturing to ensure that they would fit into the team environment. Those hired were given minimal training in the old accounting systems. Instead, they spent time working with management to satisfy its need for useful information.
Although some old accounting staff transferred successfully to the new team, others left voluntarily. In addition to resenting the “hot shot” MBAs, accountants who left were burnt out from years of fighting the complexity of the old systems. They had also spent much of their professional lives learning how these systems worked. Not surprisingly, they believed innovation threatened their competitive advantage—which was a detailed knowledge of the old systems.
Personnel turnover was not a deliberate goal at Portables, but it was probably inevitable given the rapid and significant change that took place. It affected personnel in other functional areas, as well. Staff were encouraged to make the transition; if they wished to leave, every effort was made to transfer them to other divisions within the company.
Removing and Replacing Obsolete Systems
Many of the existing accounting systems were obsolete: they were designed to collect data that no longer existed, or they reported information that was no longer used by management. It was possible to remove these systems and eliminate the cost of maintaining them. In some cases, they were replaced with simple approaches that worked well in the new environment.
Continuous Flow.
Just-in-time production is designed to “flow like a river.” This image is particularly appropriate at Portables, where families of instruments are produced on a single line. It is different, however, from the old manufacturing environment, where instruments were produced in batches.
The accounting system collected information on labor, material, and overhead by work order—the traditional focus for cost collection in a batch environment where production consists of limited runs for each of several different products. At Portables, however, work orders generated unnecessary paperwork and errors as manufacturing attempted to assign identical products to different jobs.
In lieu of work orders, the focal point for costing is now the output of a production line. A standard cost per instrument is multiplied by the number of instruments produced per day or per week. Work-in-process inventory, which is constant, is costed only when there is a change in the process or the product. Because the level of work-in-process inventory is kept low, it is easy to verify by periodic counts. Output costing fits the smooth flow of JIT production.
People Involvement.
Success in a continuous improvement environment requires that workers function as a team. They are trained in multiple functions and learn group problem-solving skills to enable them to achieve team goals. They are evaluated on their contribution to the success of the team as well as on factors such as teamwork, flexibility, breadth of skills, understanding of the business, problem solving, and reliability. As a result, the focus of productivity improvement shifts from the individual worker’s performance to the overall process’s effectiveness.
Because of this shift in focus, Portables has chosen to replace detailed reporting of direct labor time for each operation with a system of exception reporting. Direct labor workers record their time only if they are not working on normal production tasks. A worker temporarily assigned to engineering, for example, will record the time spent on this task. He or she will not record the time spent on the production line. This change is consistent with the people-involvement approach, and it has also substantially reduced the number of transactions. On one major instrument line, for example, the volume of monthly labor measurement transactions dropped from 30,000 to about 50.
Continuous Process Improvement.
The continuous improvement ethos requires a focus on all aspects of manufacturing including cost, quality, and delivery. Standard cost and budget measures, in contrast, focus exclusively on cost and consider neither quality nor delivery. In addition, standards and budgets at Portables were set once a year and so provided little guidance on a day-to-day or week-to-week basis.
Today, standards and budgets are still used for planning and for inventory valuation, but they are rarely used to generate continuous improvement.
Elimination of Waste.
Scrap and rework, once considered a normal part of production at Portables, used to be tracked and reported by the accounting department. Under the new ethos of continuous improvement, it is considered important to do work right the first time and to stop production if a defect is found. The amount of scrap and rework has dropped so dramatically that the tracking system became unnecessary and was eliminated. Responsibility for elimination of waste is given to those on the line who are performing the work.
Elimination of Inventory.
Inventory is considered unnecessary in JIT production, and strenuous efforts are made to eliminate it. A kanban system pulls inventory through the plant only when it is needed. With much of the inventory gone, Portables’ extensive inventory-tracking system became unnecessary. A new system maintains only one inventory center per production line.
Elimination of Unevenness.
JIT production should be steady; sudden changes of direction to expedite orders or to correct problems are avoided. At Portables this steadiness resulted in a constant level of work-in-process.
Once the labor, material, and overhead required for this constant inventory was determined, the actual cost of producing each instrument no longer had to be tracked. Now the standard labor, material, and overhead costs are charged direcdy to completed units, and the standard material cost is “post-deducted” from the materials inventory.
Visual Control.
Visual control is used to manage day-to-day production. This technique is possible because the line is clearly and simply laid out. Imbalances and other problems are immediately evident. Simple techniques, such as attaching clothespins to parts to mark the daily consumption of inventory, are used. (The clothespins were used originally to show line workers how little of the buffer inventory they were using and to gaintheir acquiescence in its removal.)
Monthly performance reports containing variances from standard had been of little value and are not used to support continuous improvement. Problems that might appear on performance reports should be visible when they occur and should be corrected at that time. Variances represented the “what went wrong last month” approach to control—typically, no one remembered and no one cared.
Eliminating Barriers to Continuous Improvement
Accounting systems that were designed for the 1970s may impede progress in the 1980s. Such systems now represent barriers to changing production, engineering, and marketing behavior. They are also likely to slow down the business’s adaptation to new competitive conditions. The 1970s’ accounting systems at Portables created such barriers; they are described below.
Bloated Overhead
Manufacturing overhead (particularly support over-head) was a major expense—and an obvious target for continuous improvement. Purchasing, inventory control, scheduling, expediting, supervision, engineering, and accounting itself all contributed to the spending. The cause of this over-head was the enormous number of transactions being generated by the complexity of the production process prior to the shift to continuous improvement.2
The 1970s’ accounting system, however, directed attention away from these costs. Labor-based over-head rates of 300 to 1,000 percent gave the false impression that there was a causal relationship between the incurrence of direct labor and overhead cost, and that labor was very expensive. Reacting to these rates, management focused its attention on reducing the direct labor content of the products as a means of cost control.3
A major effort of product engineering, for example, was to redesign the products to reduce direct labor costs. Engineering thought this would have a major impact on product cost via the labor-based burden rate. Each new engineering change order, however, triggered nineteen events, including design approval, vendor selection, bill of material update, and redrawing of schematics. Overtime, the repeated redesign of the product resulted in an increase in overhead that more than offset any labor savings.4
Accounting itself contributed unwittingly to the bloated overhead with systems for transaction and asset control that focused on small segments of the production process. Before the elimination of detailed labor reporting, direct labor was reported for each of forty to sixty operations per production line. Direct labor represented 3 to 7 percent of manufacturing cost, and recording increments of one-fortieth to one-sixtieth of this amount probably cost more to record, process, correct, and analyze than the actual value of the labor per operation.
Costly Engineering
Portables’ products incorporated state of the art technology and were loaded with engineering features. All this engineering was done with little regard for its impact on the cost of production, the time to market, or the customer.
Many components, for example, were used exclusively in one instrument. Some of these contained proprietary technology necessary to establish market leadership, but in many cases they were functionally identical to other, commonly used parts. The unique components cost more than the off-the-shelf components because they required separate engineering, procurement, material handling, material reliability control, and data handling. They also made the product more difficult to produce.
The additional cost of unique components was not fully reflected in the standard cost of the product. Overhead cost was proportionate to the direct labor cost of the product, which was little affected by use of unique or off-the-shelf components. Since there was a minimal cost penalty, it is not surprising that engineers favored unique components.
With a new focus on manufacturability and customer value, the engineers sought better cost information. They set up a separate product-costing system called the “Phafenwacker,” named for a mythical medieval magician. Despite this attempt to weave a little medieval magic, the effort was in vain. Accounting had a monopoly on product cost data, and an improvement in overhead costing had to wait for the implementation of a new overhead allocation system.
Building for Inventory
Maintaining a large inventory was inconsistent with the drive to continuous improvement. In 1983, most of the floor space in the plant was taken up with work-in-process inventory. There were so many parts and completed instruments that it was necessary to rent extra warehouse space. Even with all this inventory, Portables still could not get the product to the customer fast enough, because long production runs tended to produce large inventories of the wrong product. Parts shortages were frequent because of disruptions to expedite orders and to rework defective products. Building for inventory also generated a lot of defective product and scrap. It took so long for a part or subassembly to work its way through the buffer inventories that a defective part could go unnoticed for weeks. In the meantime, production of defective units continued until the lot was completed.
A major culprit in the drive to build inventory was the use of labor standards to evaluate the efficiency of direct labor and to compensate performance. This use created a direct incentive for labor to build inventory—after all, the more they built, the more favorable their efficiency variance and the higher their compensation. Plant management was happy because a high level of output ensured that overhead would be fully absorbed. It didn’t matter that there might not be a customer for the output, or that the output might be defective.
Customer Giveaways
Portables had introduced low-volume specialty products as a way of building volume and market share. The company also offered accessories as a way of tailoring the product to the exact customer need. It was assumed that these low-volume and tailored products were profitable because their reported standard cost was no different than that of high-volume, mainline products. The standard cost was based on the assumption that the consumption of overhead was proportionate to the amount of direct labor required to build the product.
In reality, the low-volume and tailored products consumed significantly more support services per unit than did the high-volume, mainline products. They increased the complexity of procurement, production, and distribution. The proliferation of these products, therefore, had a major impact on overhead cost and impeded continuous improvement.5 Table 2 summarizes the shortcomings of direct labor-oriented accounting.
Generating Continuous Improvement
Accounting measurement at Portables is meant to be proactive in the search for continuous improvement. Accounting has developed new measurements that reinforce and direct this search.
Measures of Continuous Improvement
Portables has developed several measures of manufacturing performance that report on how well manufacturing is delivering value to its customers. They include both financial and nonfinancial measures of efficiency, quality, and delivery. Continuous improvement in these variables over time is the key to achieving competitive advantage, so Portables monitors the measures for long-run trends in performance rather than for month-to-month variances from standard.
Output Rate.
The output rate is a measure of total process efficiency. It is the number of instruments produced by a production line during a day or a week. A running tally is posted above the end of the line so the production team can monitor its own performance. Management compares the output rate with the line’s production quota and looks for problems that extend beyond daily or weekly line balancing.
Output per Person.
Output per person is calculated as the cost of output achieved per person per day. The measure is calculated for all manufacturing personnel, including indirect as well as direct labor. It is used to gauge the impact of problem-solving programs, cross training, and other measures aimed at improving the productivity of the manufacturing team. It is an average measure and is not used to evaluate individual performance.
Output per Salary Dollar.
The output per salary dollar is an average for indirect and direct salaries combined. It provides a benchmark for assessing the impact on productivity of changes in the mix of personnel away from direct labor to process engineers.
Cost of Sales Ratio.
Cost of sales as a percentage of gross sales provides a measure of productivity relative to the market value of output. Portables tracks this measure over time and expects a price-adjusted downward trend. This measure is considered too gross to be used by manufacturing management and is used primarily by the division’s general manager as an indicator of competitive advantage gained by cost reduction.
Floor Space.
Floor space is a key indicator of manufacturing excellence at Portables. It is a proxy for inventory levels and for the ability to improve production flow. This measure is not calculated quantitatively, but rather assessed visually. As long as empty space continues to appear, manufacturing is making progress in eliminating inventory and improving production flow.
Cycle Time.
Cycle time is the elapsed time from the start of instrument assembly to its completion. Management uses this as a measure of efficiency. A lower cycle time reflects success in reducing inventory by reducing buffers and safety stocks, in eliminating the need for testing equipment by product or process redesign, in balancing people and equipment, in automating, and in reducing the use of space and distance traveled by the product. It is also used as a proxy for inventory turnover.
Cycle Time Efficiency.
Cycle time efficiency is direct labor time divided by cycle time. Management uses this measure to drive down the factors that cause cycle time to exceed direct labor time: waiting time (such as in queues for environmental testing), travel time from one operation to another, time spent sitting in safety stocks and buffers, and time spent in parts handling and movement.
Pass Rate.
This measure is the percentage of output that meets product specifications. It is calculated at the end of each key process to monitor the status of the process and to trigger corrective action. Pass rate is not only a key measure of quality for management; it is also used as a proxy for cost. With a higher pass rate, rework will be less, labor cost will be less, and inventory levels will be lower.
Field Failure Percent.
The reliability of the product in the hands of the customer, an important component of product quality, is measured by the percentage of failures in the field. Management watches this measure closely because field failures provide clues for improving quality.
Service Level.
Service level is how often manufacturing meets the commitment date given to the customer. Using the customer’s commitment date rather than the production schedule reflects the importance of on-time delivery in this business.
Overhead Allocation
Portables now uses overhead allocation as a means of influencing engineering decisions. High labor-based overhead rates encouraged product engineers to focus on labor-reducing changes to the product that actually made it more difficult and more costly to improve manufacturing. A new overhead method is intended to shift attention to changes that would facilitate and accelerate the improvement of manufacturing.
The overhead cost associated with procuring and carrying a part (approximately half of total manufacturing overhead) is allocated using a part-number method. The method assigns an equal overhead cost to each part number. This cost is then divided by the total volume of usage of each part number to yield a cost per part. The cost per part of a high-volume part number is thus much lower than the cost per part of a low-volume item.
This method is intended to highlight the relationship between new product design and the cost of overhead. It significantly shifts product cost from products with high-volume common parts to those with low-volume unique parts. It thus gives product engineering a direct incentive to weigh the need for a custom-designed part against the extra over-head cost required to procure and carry that part.
This method yields a cost that better measures a product’s consumption of overhead than does the old direct labor approach. Its accuracy is limited, however: it assumes that this type of overhead is incurred only as a function of the number of parts. In reality, some of the overhead cost may bear a stronger relationship to other variables, such as the number of purchase orders or the number of parts shipments received.
The method is significant in that it addressed the proliferation of part numbers at Portables. This proliferation served to increase not only overhead cost; it also increased the logistical complexity of production and procurement, and precluded the establishment of close relationships with a small number of vendors. Part-number allocation has resulted in the design of new products with a higher number of common parts and the discontinuance or repricing of older products with large numbers of custom parts.
A New Opportunity
The innovations introduced by Portables demonstrate that accounting can facilitate and drive efforts to achieve continuous improvement. Systems have been simplified to eliminate unnecessary complexity. Accounting systems and measures that were barriers to continuous improvement have been removed. New performance measures have helped manufacturing focus on key performance factors. A new allocation method has influenced product design decisions, encouraging a simpler product that is less costly and easier to manufacture.
The most important part of this story is that these changes were proposed and implemented successfully. While competition from the Japanese precipitated the changes, the creation of an “island of innovation” was the key success factor. This island included a new management team from outside the company, geographical isolation from the rest of the company, an innovative, management-oriented accounting staff, and an emphasis on team-work, participative management, leadership, and trust.
This successful accounting innovation, and the important role this innovation played in the turn-around of the division, provide an important lesson. Accounting is not by nature moribund, nor must it play a neutral or adversarial role in the management of a manufacturing business. Today’s accounting is an essential ingredient in the achievement of continuous improvement.
References
1. See H.T. Johnson and R.S. Kaplan, Relevance Lost: The Rise and Fall of Management Accounting (Boston: Harvard Business School Press, 1987).
2. For a description of the impact of transactions on overhead, see J.G. Miller and T.E. Vollmann, “The Hidden Factory,” Harvard Business Review, September–October 1985, pp. 142–150.
3. This problem was solved by the parts burdening system described later in the paper.
4. An engineering change order is an example of what Miller and Vollmann call a change transaction in “The Hidden Factory,” p. 146.
5. For a discussion of the impact of product proliferation on complexity, and the failure of traditional cost systems to reflect the cost of this complexity, see R. Cooper and R.S. Kaplan, “How Cost Accounting Systematically Distorts Product Costs,” in Field Studies in Management Accounting and Control, W. Bruns and R.S. Kaplan, eds. (Boston: Harvard Business School Press, 1987).