The Universal Motor Company is one of the world’s largest manufacturers of automobiles and trucks. In 1980, purchase content was 65 percent of the cost of goods sold. The supply management organization was well managed and staffed with seasoned, well-educated professionals (for the time period).
During the late 1970s, the U.S. government began addressing the air pol1ution issue by establishing vehicle emissions standards and corporate average fuel consumption targets for vehicles sold in the United States. The auto industry was faced with major technical problems. The U.S. government-mandated standards could not be achieved with available technology. These mandated standards required the auto industry to design engine control computers (fuel intake, spark timing, etc.) to manage engine efficiency more precisely.
Management was highly confident that established supply management policies and procedures that had been used successfully for so many years would apply to the procurement of the new semiconductors required for the manufacture of the new engine control computers. By 1980, it was obvious to the individuals directly involved that something was wrong. The tried and true methods of supplying stampings and plastic injection molded parts were not getting satisfactory results when applied to the procurement of semiconductors. The company purchased about $10 million worth of semiconductors in 1980 from twenty suppliers. Stockouts leading to production disruptions (absolutely forbidden in the auto industry) were occurring regularly. In fact, semiconductors were a greater cause of production disruptions than were all other purchased materials. Semiconductor suppliers, for the most part, seemed uninterested in the firm’s problems and did not react to these emergencies in the way that more traditional auto industry suppliers react.
Those responsible for the procurement of semiconductors were under intense pressure from management to resolve these difficulties. But nothing they tried seemed to make a difference. During these dark days some important observations were made:
• The total 1980 worldwide auto industry semiconductor requirement represented less than 2 percent of the semiconductor market. Therefore, members of the semiconductor industry did not see the auto industry as an important market requiring or deserving any special service.
• The supply management practice of sourcing from many suppliers (to ensure competition and a low purchase price) was aggravating an already difficult supplier relations issue and further complicating the situation.
• Supply management projected its semiconductor requirements out over the next five years. The projection was shocking: The current $10 million annual requirement was forecast to grow to over $90 million by 1985. It seemed likely that this growth would be replicated by the entire auto industry. However, this growth did not increase the market position of the auto industry, as the semiconductor industry was growing faster than the automotive industry market component.
Based on these observations, it became clear that something had to change. The 1980 semiconductor supply situation was intolerable. Of equal or greater concern, projected growth in electronics requirements appeared to be unsupportable.
In 1980, the concept of forward planning, business planning, or strategic planning was not commonly associated with the supply function. Supply management was most often thought of as a reactive function organized to support manufacturing. Marketing, on the other hand, was more often thought of as the proactive organization focused on the future of the firm. The failure of management to recognize the strategic and futuristic implementation of supply management limited innovative actions when dealing with a firm’s supply world.
The 1980 situation demanded new thinking and a new approach to the acquisition of semiconductors. The supply managers involved had little choice, given the circumstances. Supply had no experience with strategic planning. There were no guidelines to follow. The responsible supply managers started with a clean piece of paper, a mission, and a problem statement.
The first order of business was to develop a complete understanding of both the internal and external environments. The firm’s engineers documented their future technology requirements in the form of a technology plan or road map. Once the internal requirements were understood and documented, the focus turned to the external environment, revealing an industry which was (and remains) dynamic, high growth, entrepreneurially focused, and technology driven—one that defies simple analysis.
Areas of Strategic Importance
Several major areas were identified as being of strategic importance to the firm’s survival and future profitability: How can Universal ensure that it benefits from rapidly evolving technologies in the area? How can Universal obtain the desired/required quality? How can Universal best ensure a continuity of supply? What actions need to be taken to compress the time involved in going from concept to customer? Can manufacturing cycle time be compressed? How can Universal best minimize the “all-in-cost” associated with the acquisition and use of semiconductors?
The Make-or-Buy Issue
One of the major issues confronting the team had to do with whether the firm should make or buy the required semiconductor devices. (We will explore this issue in considerable detail since it is representative of the complexity of decisions in the areas of design, source selection, pricing, and related supply management issues. Further, the make-or-buy decision impacts on all six of the areas of strategic importance.)
In order to understand the make-or-buy question, management recognized the need for a complete environmental analysis of the semiconductor industry. A team of company experts including engineering, quality, supply, and manufacturing professionals was assigned to develop a program analyzing the semiconductor industry. Both primary and secondary researches were conducted. The research led Universal executives to conclude that the decision was more
complicated than simply deciding whether to acquire a semiconductor facility or buy the required components. It became apparent that the semiconductor make-or-buy issue consisted of many subissues. Accordingly, the entire process of designing, manufacturing, and testing semiconductors was broken down into stages.
The first stage (design) involved both human and computerized inputs. Computerized design was in its infancy. Carver Meade had recently written his book addressing cell-based design methodology. There was a CAD/CAM component to design. But most design was done without a lot of computer-aided blocks or cells by human design engineers.
The second phase consisted of the development of masks for the process. The third phase had to do with the actual manufacture of the semiconductor, including the manufacture of the silicon ingot and imprinting of the integrated circuits. The fourth phase involved testing the manufactured wafers. The fifth phase involved packaging the integrated circuits. The last phase had to do with testing the packaged circuits. Thus, detailed investigation revealed that there was a whole array of different manufacturing processes and engineering technologies involved and that the make-or-buy issue was not quite as simple as was originally thought.
1982 Environment Scan
A 1982 environment scan revealed some very interesting facts. The first and most important was that design engineers were in incredibly short supply. There were fewer semiconductor design engineers in the world than there were professional football players in the NFL.
A second finding was that there was adequate mask-making and manufacturing capacity for making ingots and imprinting integrated circuits. Manufacturing capacity (equipment and clean rooms) is incredibly expensive. To exacerbate the situation, technology moves very quickly and the obsolescence rate of the required capital equipment is quite high. For these reasons, senior executives properly were concerned about manufacturing capacity. But the research indicated that there was adequate capacity available worldwide and that sufficient capacity was being added to meet future demand.
The research effort revealed that the packaging issue consisted of at least two subissues: industry capacity and packaging technology. Capacity did not appear to be a problem. But management had some concerns about packaging technology to which the semiconductor industry was not sensitive. The semiconductor industry’s attitude toward packaging was that parts would be placed into standard packages that would be good for all of the industry to use. The semiconductor industry’s attitude was in conflict with Universal’s ideas on manufacturing efficiency and module packaging, broad layout, size, weight, and other design requirements. Universal’s ideas would allow the production of customized packages for special opportunities on new models.
The issue of testing of the packaged semiconductor devices was extremely critical to Universal. The issue had to do with the relative value of a semiconductor to the various suppliers and to Universal. A semiconductor sold for $1 to $3, while an automobile, dependent on the semiconductor’s performance, sold for $10,000 to $30,000 (1980 dollars). A $1 to $3 component could easily (and all too frequently did) cause a $20,000 automobile to malfunction, requiring the owner of the automobile to call a tow truck, pay for the tow truck, and have the vehicle towed in
for repairs costing several hundred dollars. This is not the basis of customer satisfaction or of increasing market share for the auto manufacturer. The importance of defect-free semiconductors was of far greater importance to Universal than to the semiconductor industry, whose liability was limited to replacing the defective part. Universal executives recognized that the semiconductor industry did not have the same motivation to quality as it had.
After the above analysis was completed, another issue influencing the make-or-buy issue came into focus: technology leadership. A chip might offer Universal technological leadership in the automobile industry and possibly some very important marketing advantages. Management became very nervous about its semiconductor supplier(s) being able to sell identical or similar chips to the firm’s competitors. Universal wanted to own the technologies incorporated in these circuits. These technologies were viewed as strategic keys to Universal’s future.
Management, with the assistance of legal council, determined that if Universal designed the chips in-house, for all practical purposes it would own the technology. The processing technology would be covered by contract license agreements.
The economics involved in the make-or-buy issue were unusually complex. In 1982, the manufacturing facilities that were required to make a meaningful volume of semiconductors for Universal would have cost in excess of $50 million. There was concern about Universal’s ability to run a high-tech semiconductor facility effectively and efficiently. Further, there was the issue of culture. Would the auto industry culture and the semiconductor industry culture mesh? It became apparent that a decision to manufacture integrated circuits (ICs) was a very high cost strategy. It was likely that the resulting ICs (manufactured by Universal) would cost two to four times as much as purchased ones. However, management felt that the cost penalty might be a sound investment in quality and technology and give Universal a significant competitive advantage.
What alternatives are available to Universal and explain advantages and disadvantages of each?