Which of the following may be scheduled in production planning by the use of learning curves?
Many companies have built successful marketing and production strategies around the learning curve—the simple but powerful concept that product costs decline systematically by a common percentage each time that volume doubles. The learning-curve relationship is important in planning because it means that increasing a company’s product volume and market share will also bring cost advantages over the competition. However, other results that are not planned, foreseen, or desired may grow out of such a market penetration cost reduction progression. Reduced flexibility, a loss of innovative capability, and higher overhead may accompany efforts to cut costs. A manager failing to consider the possible outcome of following a cost-minimizing strategy may find himself with few competitive options once he reaches the point where decelerating volume expansion prevents him from obtaining further significant cost reduction. But if he can identify the likely consequences in advance, he can either anticipate them in his plans or choose an alternative strategy. In this article we analyze those consequences and conclude that management cannot expect to receive the benefits of cost reduction provided by a steep learning-curve projection and at the same time expect to accomplish rapid rates of product innovation and improvement in product performance. Managers should realize that the two achievements are the fruits of different strategies. Proponents of the learning curve have developed the relationships between volume growth and cost reduction through the use of two distinct but related approaches: 1. The learning curve (also called the progress function and start-up function) shows that manufacturing costs fall as volume rises. It has typically been developed for standardized products like airframes and cameras. 2. The experience curve traces declines in the total costs of a product line over extended periods of time as volume grows. Typically, it includes a broader range of costs that are expected to drop than does the learning curve, but disregards any product or process design changes introduced during the period of consideration. Gas ranges and facial tissues are two major product lines on which experience curves have been developed. The two approaches are sufficiently similar for many purposes of planning and analysis. As we shall demonstrate in due course, however, changes in pricing policy and product design can create significant discrepancies. Care must be exercised in choosing between the two related approaches. Hard Strategic QuestionsEvidence on cost decreases in a wide range of products, including semiconductors, petrochemicals, automobiles, and synthetic fibers, supports the notion that total product costs, as well as manufacturing costs, decline by a constant and predictable percentage each time volume doubles. Because this volume/cost relationship is reliable and quantifiable, it has appeal as a strategic planning tool for use in marketing and financial planning, as well as in production. Moreover, a strategy that seeks the largest possible market share at the earliest possible date can gain not only market penetration but also advantages over competitors who have failed to reach equal volume. Examples of the economic effects of the learning curve can be found everywhere. The price of ferromagnetic memory cores for computers plunged from 5 cents per bit (unit of memory) in 1965 to less than a half cent in 1973, thereby significantly reducing the costs of computers. In less than two decades of production DuPont reduced the cost of rayon fiber from 53 cents a pound to 17 cents (values not adjusted for inflation). Airframe costs can drop more than 50% per pound during the three to five years of a high-volume production run if the manufacturer can control the rate of modification and sustain volume production. In considering examples of independent action by one corporation, the most important is that of the Ford Motor Company in its early years. (The Ford example actually shows an experience curve, but the point it makes is equally valid for a learning-curve situation.) During an initial period of less than two years, the average price of a Ford automobile was reduced from more than $5,000 to about $3,000 through the introduction of a dominant product, the Model T. Then, as Exhibit I shows on a logarithmic scale, the company cut the price of the Model T to less than $900 following an 85% experience curve. (To underline the contrasts in price, all the figures are translated into 1958 dollars.)  Exhibit I Price of Model T. 1909–1923 (Average List Price in 1958 Dollars) During this time span wages were increased more than threefold, the working day reduced by fiat from ten hours to eight, the moving assembly line invented, and one of the nation’s largest industrial complexes (River Rouge) created entirely out of retained earnings. We shall return to the Ford case shortly. The frequency with which this cost reduction/ volume increase pattern is found in practice sometimes leads to the incorrect impression that the learning-curve effect just happens. On the contrary, product design, marketing, purchasing, engineering, and manufacturing must be carefully coordinated and managed. The producer cuts costs with a combination of effects; these include spreading overhead over larger volume, reducing inventory costs as the process becomes more rational and throughput time drops, cutting labor costs with process improvements, achieving greater division of labor, and improving efficiency through greater familiarity with the process on the part of the work force and management. The impetus toward lower costs and higher volume is fragile, however, and if any one of the necessary conditions is removed, a discontinuous return to higher costs may result. The question management must ask in undertaking such a strategy is whether it fully anticipates or desires the implications that accompany results or that follow execution of the strategy. After the startup phase, doubling of volume has tremendous implications for the organization. Not all the changes it undergoes may be desirable. Management must anticipate the consequences so that it can plan for them, or else it should reject the strategy from the beginning. Some of the questions that it must ask itself are:
To explore these questions, we shall consider Ford’s early experience, particularly with the Model T. Then we shall examine other manufacturing cases—such as TV picture tubes, electronic components, and office equipment. The evidence suggests that with those products whose performance can be improved significantly—typically involving complex manufacturing processes such as use of electronic equipment machinery—the incidence of product innovation establishes the limit to the learning curve. The consequence of intensively pursuing a cost-minimization strategy is a reduced ability to make innovative changes and to respond to those introduced by competitors—although the amount of loss seems to depend on the degree to which the manufacturer follows such a strategy, and its intensity. The problem of strategy choice, then, is balancing the hoped-for advantages from varying degrees of cost reduction against a consequent loss in flexibility and ability to innovate. From Model T to Model AAt Ford, the experience curve did not continue indefinitely; it governed only the Model T era. Then Ford abandoned it for a performance-maximizing strategy by which the company tried to improve performance year by year at an ever higher product price. The product was the Model A. However, Ford’s long devotion to the experience-curve strategy made the transition to another strategy difficult and very costly. Exhibit II shows volume and average prices of the Ford line for some 60 years in an experience-curve format. (The scale of the top part is chronological; the bottom part is logarithmic.) Data on retail price trends, displayed by the two curves, are related to both product-line diversity and the rate of product change. Data on the variety of wheel bases and engines, the horsepower range offered, and the average vehicle weight illustrate how the number of options expanded, contracted, and expanded again. An indicator of the changes in models appears at the top of the exhibit. Taking these three types of information together—product line diversity, the rate of model change, and price trends—one can see that they changed concurrently, whether price is defined on a per-vehicle basis (the upper trend line) or on a per-pound basis (the lower). Exhibit II The Ford Experience Curve (in 1958 Constant Dollars) Because manufacturing costs vary directly with weight, a comparison of the two trend lines in different periods is revealing. After the Model T was discontinued in 1927, Ford raised the price of its car from year to year, in contrast to the earlier period. The increases were due mainly to design changes which were made to enhance comfort, performance, and safety, but which required more and more expensive materials and caused the price per pound to rise steadily. Considered over a number of years, these systematic annual changes represent a tradeoff in favor of size, weight, and performance, as opposed to price. As the exhibit shows, after an initial period in which several models were offered at the same time, the product line was consolidated in 1909 to the Model T. Ford’s objective was to reduce the price of the automobile and thereby increase volume and market share. Before the Model T was conceived, when the least expensive Ford car was priced at $850 and tires alone cost more than $60 a set, Henry Ford announced plans to sell autos at $400—although, he told reporters, “It will take some time to figure what we can do.” By 1907, after the death of the former company president and the expulsion of dissident stockholder-managers who advocated high-priced cars, attention turned to product cost reduction. The company felt confident in taking this step because of its success with the relatively inexpensive Model N in 1907 and later with the Model T, which was clearly a superior product.1 The company accomplished savings by building modern plants, extracting higher volume from the existing plant, obtaining economies in purchased parts, and gaining efficiency through greater division of labor. By 1913 these efforts had reduced production throughput times from 21 days to 14. Later, production was speeded further through major process innovations like the moving assembly line in motors and radiators and branch assembly plants. At times, however, labor turnover reportedly ran as high as 40% per month.2 Up to this point, Ford had achieved economies without greatly increasing the rate of capital intensity. To sustain the cost cuts, however, the company embarked on a policy of backward and further forward integration in order to reduce transportation and raw materials costs, improve reliability of supply sources, and control dealer performance. The rate of capital investment showed substantial increases after 1913, rising from 11 cents per sales dollar that year to 22 cents by 1921. The new facilities that were built or acquired included blast furnaces, logging operations and saw mills, a railroad, weaving mills, coke ovens, a paper mill, a glass plant, and a cement plant. Throughput time was slashed to four days3 and the inventory level cut in half, despite the addition of large raw materials inventories. The labor hours required of unsalaried employees per 1,000 pounds of vehicle delivered fell correspondingly some 60% during this period, in spite of the additions to the labor force resulting from the backward integration thrust and in spite of substantial use of Ford employees in factory construction. Constant improvements in the production process made it more integrated, more mechanized, and increasingly paced by conveyors. Consequently, the company felt less need for management in planning and control activities. The percentage of salaried workers was cut from nearly 5% of total employment for 1913 to less than 2% by 1921; these reductions in Ford personnel enabled the company to hold in line the burgeoning fixed-cost and overhead burden. The strategy of cost minimization single-mindedly followed with the Model T was a spectacular success. But the changes that accompanied it carried the seeds of trouble that affected the organization’s ability to vary its product, alter its cost structure, and continue to innovate. Cost of transitionIn its effort to keep reducing Model T costs while wages were rising, Ford continued to invest heavily in plant, property, and equipment. These facilities even included coal mines, rubber plantations, and forestry operations (to provide wooden car parts). By 1926, nearly 33 cents in such assets backed each dollar of sales, up from 20 cents just four years earlier, thereby increasing fixed costs and raising the break-even point. In the meantime, the market was changing. In the early 1920s, consumer demand began shifting to a heavier, closed body and to more comfort. Ford’s chief rival, General Motors, quickly responded to this shift with new designs. Ford’s response was to add features to the Model T which gradually increased the weight; between 1915 and 1925 the weight of the car actually gained by nearly 25%, while engine power remained the same. But the rate of product improvement halted the steady reduction of costs. Nevertheless, to maintain market growth Ford further cut the list price along the experience-curve formula. This created a severe margin squeeze, particularly when unit sales began falling after 1923. As the rate of design changes accelerated and wage levels continued to rise, manufacturing costs loomed ever larger in the retail price. In 1926, the manufacturing costs of some models reached 93% of list price, and some models were actually sold to dealers at prices below costs. (See Exhibit III for sales, manufacturing, and other data on Ford during the critical two decades.) Ford, unbeatable at making one product efficiently, was vulnerable to GM’s strategy of quality and competition via superior vehicle performance. As Alfred Sloan, architect of GM’s strategy, later wrote: Exhibit III Ford Vital Statistics, 1910–1931 Sources: Ford Archives; Federal Trade Commission, Report on the Motor Vehicle Industry, 76th Congress, First Session (1940), House Document 468. Missing figures are not available. “Mr. Ford…had frozen his policy in the Model T,…preeminently an open-car design. With its light chassis, it was unsuited to the heavier closed body, and so in less than two years [by 1923] the closed body made the already obsolescing design of the Model T noncompetitive as an engineering design… “The old [GM] strategic plan of 1921 was vindicated to a ‘T,’ so to speak, but in a surprising way as to the particulars. The old master had failed to master change… His precious volume, which was the foundation of his position, was fast disappearing. He could not continue losing sales and maintain his profits. And so, for engineering and market reasons, the Model T fell… In May 1927…he shut down his great River Rouge plant completely and kept it shut down for nearly a year to retool, leaving the field to Chevrolet unopposed and opening it up for Mr. Chrysler’s Plymouth. Mr. Ford regained sales leadership again in 1929, 1930, and 1935, but, speaking in terms of generalities, he had lost the lead to General Motors.”4 A company that had developed and introduced eight new models during a four-year period, before undertaking the cost-minimization strategy, had subsequently so specialized its work force, process technology, and management that it consumed nearly a year in model development and changeover. As an illustration of its specialization, in the course of the model change Ford lost $200 million, replaced 15,000 machine tools and rebuilt 25,000 more, and laid off 60,000 workers in Detroit alone. So we see that when costs could not be reduced as fast as they were added through design changes, the experience-curve formula became inoperative. While this sequence should give pause to managers who wish to apply the experience curve to make product-line changes, it does not invalidate the principle of the learning curve, which assumes a standardized product. Decline of InnovationThe sequence of evolutionary development in product and process during the period of the cost-minimization strategy and the subsequent strategy transition is paralleled in the pattern of major Ford innovations. Exhibit IV plots the frequency and significance of Ford-initiated innovations by type of application: product innovation, process innovation, and transfer of process technology to or from associated industries. The new methods and designs are those claimed by Ford. For our analysis, four independent industry experts evaluated the importance of each one and rated it on a scale of 1 to 5. The innovations range in significance from the introduction of the plastic steering wheel (index average of 1) in 1921 to the invention of the power-driven final assembly line (index of 5) in 1914. The vertical axis in Exhibit IV provides a sum of the average points assigned to significant developments by two-year intervals in Ford’s history. Exhibit IV Innovation and Process Change at Ford The exhibit indicates that the intensity of innovative activity is closely related to major events in the unfolding of the cost-minimization strategy. During the Model T period the activity shows a ripple effect. Installation of new product applications occurs in clusters with new model development and then declines in frequency as the design is standardized, efficiency is refined, and the process is integrated into operations. Process innovations rise to a peak after the period of product innovation, as the manufacturer rationalizes the process and reduces costs. (Compare the peak designated circled 1 with the peak designated squared 1, circled 2 with squared 2, and so on.) As the manufacturer works out these problems, he transfers process technology following the thrust into backward integration, and a third peak of activity occurs (triangled 2, triangled 3, and so on). The exhibit suggests not only that the nature of innovation changes, but also that the intensity of innovative activity diminishes. Ford produced only one new product application or process technique during the seven years after 1932 that rated as high on the scale as 4—the development of transfer machines. This step toward further automation took place in 1937. The changes introduced to trim costs altered the innovative activity in two ways. First, after 1926 the types of innovation peaked coincidentally. As operations became more elaborate and systemslike, product and process change developed intimate linkages; many different elements had to be altered simultaneously to introduce change. This relationship implies a high cost of change. Secondly, the nature of product innovation shifted. In the early years, a new model meant a complete transformation involving major innovation. Later, model change became an annual affair, and innovation centered on new features available across model lines rather than on new models. For instance, the V-8 engine, whose development appears as a substantial cluster of innovations in Exhibit IV, was produced without substantial alterations for 18 years. Not surprisingly, the third class of innovation, technology transfers, increased in frequency through the period under consideration. This class had particularly long-term value at Ford since it improved the manufacturing capability. Many of these transfers were accomplished in Ford’s newly integrated feeder operations, such as one where technology was applied to produce plate glass continuously. Ford’s experience demonstrates the important link between innovation and strategy. Innovation is not the pacing element; it is part of the strategy. Ford’s choice of strategy made innovation more costly and a more serious organizational problem. Unfortunately, the cost-cutting drives also led to weakening of the resources (the salaried employees) needed to initiate and carry out innovation. It is not surprising that the company took nearly a year to change over to the Model A. With its new model, Ford rose again. Combining the old philosophy of cost reduction with the appeal of an entirely new car boasting demonstrably high performance, the company wrestled the major market share from GM in 1930. But its market share fell once more. Indeed, Chrysler, a distinct third among auto makers during the 1920s, held second place ahead of Ford during most of the Depression. As it turned out, the company’s highly specialized production process lacked the balance to handle the new product; for example, the company had overcapacity in wood (the Model T had many wooden parts) but undercapacity in glass and body parts manufacturing. Moreover, as indicated by the data in Exhibit III, Ford never regained the high levels of labor and capital productivity of its heyday. Despite extensive investments in new plant and equipment, even in the highest volume year for the Model A (1929), 40 cents in plant and equipment assets were required per dollar of sales, and nearly 80 hours of direct labor were required per vehicle. Ford did not improve on these figures until the late 1940s, when new management restructured the company and made heavy plant investments. From the time it introduced the Model A, Ford was compelled to compete on the basis of product quality and performance—a strategy in which it was not skilled. Airframes, Computers, and so onThe Ford case provides a spectacular example of one company’s action in pursuing a cost-minimization strategy to its end. Although this is an extreme case in terms of strategy choices and investment magnitudes, the same forces and consequences can be found at stake in other industries. In some cases these forces and consequences are evident when a rapid rate of product change retards the inauguration of the learning curve, and in other cases the difficulties terminate the downward trend. Consider:
The contrary relationship between product innovation and efficiency exists not only in instances where the impetus for change comes about after a long and successful production run, as in the Ford case and in that of Volkswagen more recently. It can also be found when the change is an unintended continuation of uncertainty following new model introduction, as happened in the foregoing airframe and computer examples. Common Elements of ChangeTo consider the sort of changes that can accompany a cost-minimizing strategy, it is useful to abstract that aspect of the Ford case. The kinds of changes that took place can be grouped into six categories—product, capital equipment and process technology, task characteristics and process structure, scale, material inputs, and labor. Product:Standardization increases, models change less frequently, and the product line offers less diversity. As the implementation of the strategy continues, the total contribution improves with acceptance of lower margins accompanying larger volume. Capital Equipment and Process Technology:Vertical integration expands and specialization in process equipment, machine tools, and facilities increases. The rate of capital investment rises while the flexibility of these investments declines. Task Characteristics and Process Structure:The throughput time improves and the division of labor is extended as the production process is rationalized and oriented more toward a line-flow operation. The amount of direct supervision decreases as the labor input falls. Scale:The process is segmented to take advantage of economies of scale. Facilities offering economies of scale, such as engine plants, are centralized as volume rises, while others, like assembly plants, are dispersed to trim transportation costs. Spreading the higher overhead over larger volume gains savings. Material Inputs:Through either vertical integration or capture of sources of supply, material inputs come under control. Costs are reduced by forcing suppliers to develop materials that meet process needs and by directly reducing processing costs. Labor:The heightening rationalization of the process leads to greater specialization in labor skills and may ultimately lessen workers’ pride in their jobs and concern for product quality. Process changes alter the skills requirements from the flexibility of the craftsman to the dexterity of the operative. The same pattern of change in the six categories that characterizes the Ford history also describes periods of major cost reduction in other industries. For example, as light-bulb manufacturing progressed from a manual process to an almost entirely automated one, a similar pattern of product development, process elaboration, increase in capital intensity, and so on, was evident.6 In areas as diverse as furniture manufacturing and commercial building construction, the problems of improving productivity and achieving innovation often hinge on changes similar in thrust to those at Ford. Life-cycle studies of international trade in many products, such as chemicals and petrochemicals, demonstrate a coordinated pattern of change involving product characteristics, scale, and price competition that is consistent with the Ford case. Studies of manufacturing technology yield a common finding for electronics, chemical, and metalworking companies, among others, that certain conditions in a company, like its supervisory structure, product-line diversity, and utilization of technology, relate to characteristics of the manufacturing process. More specifically, manufacturers with more efficient line flows have different ratios of supervisory personnel to the work force, different levels of authority, less product diversity, and greater product standardization than manufacturers with more flexible production process structures. Risks of SuccessIn analyzing the difficulties of Ford and other companies, we are not arguing that the pursuit of a cost-minimization strategy is inappropriate. The failure of many companies, particularly small, innovative ones, can be traced to their inability to make the transition to high volume and cost efficiency. Nevertheless, management needs to recognize that conditions stimulating innovation are different from those favoring efficient, high-volume, established operations. While there must be a theoretical limit to the amount by which costs can ultimately be reduced, a manufacturer reaches the practical limit first. However, the practical limit is not reached because he has exhausted his means of cutting costs; it is rather determined by the market’s demand for product change, the rate of technological innovation in the industry, and competitors’ ability to use product performance as the basis for competing. In determining how the learning-curve strategy should be pursued, management must realize that the risk of misjudging the limit rises directly with the successful continuation of the strategy. There are two reasons for this seemingly paradoxical development: first, the market becomes increasingly vulnerable to performance competition and second, attempts to continue reducing costs diminish the organization’s ability to respond to this kind of competition. The market becomes more vulnerable to performance competition because the company must stake out an ever-larger market share to maintain a constant, significant rate of cost cutting. Demand must be doubled each time in order to realize the same proportional cost reduction. As the market expands, it becomes harder to hold together and the competition is better able to segment it “from the top,” with a superior product or customized options. Once this action is taken, the company on the learning curve must either abandon the all-important volume bases of scale or introduce a major product improvement. Either step, or both, ends the cost-reduction sequence. The unfortunate implication is that product innovation is the enemy of cost efficiency, and vice versa. To make the learning curve evolve successfully, the manufacturer needs a standard product. Under conditions of rapid product change, he cannot slash unit output costs. Managing TechnologyThe role expected of technology is critical in the formulation of manufacturing strategy. Many a company has sailed into the unknown, trailing glowing reports about the R&D under way in its laboratories and the new products it is developing. Yet too often the promises in annual reports to stockholders and in news releases are never realized. The problem hinges on difficulties in recognizing that a shift in strategy has a pervasive effect across the organization’s functional areas. The production department cannot follow a program of cost reduction along the learning curve at the same time that R&D or the marketing people are going full steam ahead into new ventures that change the nature of the product. When a new product born of technology fails, management is often chided because it assertedly marketed the product poorly. The problem may have come, however, from management’s failure to realize that its capabilities to handle innovation had weakened. Foresight is a matter of judging the challenge in terms of altered capabilities as well as technological changes and market forces. In the Ford case the difficulties arose as much from what the organization did to itself as from GM’s actions. The ability to switch to a different strategy seems to depend on the extent to which the organization has become specialized in following one strategy and on the magnitude of change it must face. An extreme in either factor can spell trouble. Very little is known about how to plan for this type of technological change. But we can point to two courses of action that some major companies have followed in avoiding the problems we have described. One is to maintain efforts to continue development of the existing high-volume product lines. This requires setting the industry pace in periodically inaugurating major product changes while stressing cost reduction via the learning curve between model changes. This course of action—which IBM has followed in computers—is obviously a costly option which only companies with large resources should undertake. It amounts to a decision to maintain comparatively less efficient operations overall. The second course of action is to take a decentralized approach in which separate organizations or plants in the corporate framework adopt different strategies within the same line of business. Several corporations in high-technology industries have taken this approach with success. One organization in the company will pursue profits with a traditional product, like rayon, to the limit of the experience curve. At the same time a new, different organization will undertake the development of innovative (perhaps even competitive) products or processes, such as nylon. In taking this tack, some companies have shut down old plants and started up new ones instead of mingling different capabilities that are at various stages of their development.7 Neither of these courses of action will suit the needs of every organization, but some means of dealing with the issue of technological change and strategy transitions should be included in strategic planning. 1. Allan Nevins, Ford: The Times, the Man, the Company (New York, Scribner, 1954), Chapter XII. 2. Keith Sward, The Legend of Henry Ford (New York, Rinehart, 1948), p. 51. 3. See Factory Facts From Ford (Detroit, Ford Motor Company, 1924). 4. Alfred P. Sloan, Jr., My Years With General Motors (New York, Doubleday, 1964), pp. 162–163. 5. John Mecklin, “Douglas Aircraft’s Stormy Flight Plan,” Fortune, December 1966, p. 258. 6. See James R. Bright, Automation and Management (Boston, Division of Research, Harvard Business School, 1958). 7. For more on this approach, see Wickham Skinner, “The Focused Factory,” HBR May–June 1974, p. 113. A version of this article appeared in the September 1974 issue of Harvard Business Review. What is a learning curve in production?Key Takeaways. The learning curve is a visual representation of how long it takes to acquire new skills or knowledge. In business, the slope of the learning curve represents the rate in which learning new skills translates into cost savings for a company.
What are the main uses of learning curve?The learning curve model helps track training progress, improve productivity, and predict learners' performance and improvement over time. However, the relationship between the amount of time spent practicing an activity, and the overall performance is not linear.
In what way does learning curve helps management in planning control and decision making?These curves help demonstrate the cost per unit of output decreases over time with the increase in experience of the workforce. Learning curves and experience curves is extensively used by organization in production planning, cost forecasting and setting delivery schedules.
Which of the following cost is affected by learning curve?Under the impact of learning curve effect the cost per unit registers a decline and leads to movement from 'LAC 1' to 'LAC 2' indicating that on account of 'Learning Effect' less cost per unit is required to produce same quantity that is 'Q' of the Good.
|
