The Pros and Cons of Outsourcing

Total Costs and Their Various Components Must Be Scrutinized

Bijan Moslehi, PhD, is chief technology officer and senior vice president, semiconductor technology research, for The Noblemen Group, a boutique investment banking, strategic advisory, and business development firm. Moslehi has 20 years' experience working in the semiconductor and semiconductor equipment industries. He can be reached at bmoslehi@noblemengroup.com.

Outsourcing has been increasingly used by many electronics and semiconductor companies, including systems providers, chip manufacturers, and equipment suppliers. Both domestic and offshore outsourcing have given rise to a host of specialized contract manufacturing  houses and contract service providers covering virtually all aspects of the value chain. The vast and growing outsource network and infrastructure includes software development services, electronic manufacturing services, original-design manufacturers (ODMs), distributors, assembly and packaging houses, electrical test services, photomask shops, chip fabrication foundries, intellectual property (IP) providers, and design services.

Equipment maintenance, materials, and facilities management systems are among the areas of fab operation also being outsourced. Equipment suppliers also increasingly rely on outsourcing, particularly for gas panels, subassembly units, automation, and even the manufacturing of select tools. As a result, some suppliers have evolved into tool integrators, doing only final assembly and test in-house. In addition, many venture capital firms require an outsourcing plan and strategy from start-up companies that they fund.

In the 1960s and 1970s, several integrated device manufacturers (IDMs) began to move their assembly and packaging offshore in wholly owned plants, mostly located in Asia. Independent Asian test, assembly, and packaging contract manufacturers eventually formed, which in turn prompted IDMs to outsource a large portion of that work to these offshore providers.

Over the past two decades, fabless semiconductor companies have fully leveraged this burgeoning infrastructure. The emergence of foundries in the 1980s was primarily driven by the needs of fabless firms, which has greatly contributed to the success of both industrial models.  Foundries have also come to be used by the IDMs to help manage capacity, up-front capital investments, and business cycles. As a result, foundries manufacture 20–25% of all chips, while about 25% of chip assembly is outsourced. An increasing number of design jobs are being outsourced, although most of them have gone to domestic U.S. contractors. The traditional integrated product manufacturing approaches have weakened to the point of disintegration; nearly every aspect of system and chip production, from design to sales, can be outsourced.

For this issue's column, I will highlight and review certain critical aspects of outsourcing. If used properly and implemented correctly, the outsourcing option can be a powerful business tool. Two oft-cited reasons for outsourcing are flexibility and cost reduction. Outsourcing can also help leverage the expertise and domain knowledge of the outsourced partners, achieving potentially faster time-to-market and improved access to local markets.

Flexibility provides a substantial level of control over profitability during the business cycles. It does so by reducing and controlling the fixed costs of the business through the minimization or avoidance of the adverse economic effects of low factory utilization and idle factory capacity, while maintaining upward or downward scalability of the production volume. Reduced capital expenditure is another benefit of outsourcing, because of the resulting decrease in depreciation costs. Fabless companies avoid wafer fab–related capex altogether. Flexibility and the potential for little or no upfront capital spending have become the primary drivers of outsourcing. The benefits of flexibility are numerous: improved financial control through the cycles, reduced operational inefficiencies and expenses (particularly cyclically underutilized workforce and capacity during downturns), and better prospects for long-term profitability averaged over a full business cycle. A current example is the "fab-lite" strategy of several IDMs that outsource some of their manufacturing—typically 35–50%—to foundries.

Outsourcing's cost-reduction aspect recently has come under scrutiny. The popular notion that high-tech outsourcing is mainly used to reduce labor costs needs critical examination.  In many cases, the labor costs account for less than 10% of total expenses. This percentage drops to below 5% of the total for modern automated operations, such as 300-mm fabs. Since the labor force in these operations must be experienced, with as much as 40% degreed and highly skilled, a price will be paid for using cheap, "green" labor and going through a long learning curve in these critical functions. Although the relatively low cost of offshore labor is compelling, the outsourcing of many complex high-tech projects just for the sake of reduced personnel expenditures can be risky and costly.

There are alternative cost-reduction strategies. For example, materials usually account for between 50% and 80% of total expenses. Wouldn't it also be important to focus on reducing materials costs by employing value-engineered designs and sourcing lower-cost materials and components? Sometimes the real savings achieved from outsourcing are found to be greatly exaggerated, especially once the actual expenses are at hand and all the hidden costs are factored in.  Hidden costs include travel and communication, shipping, restructuring, and project management and monitoring, as well as the cost of delays, errors, and other deviations.

The disruptive impact of the outsourcing trend (particularly offshore) on the people whose jobs are weakened or eliminated has resulted in many intense discussions and heated debates in the media. The environment has become highly charged and politicized, often putting governments of the industrialized economies and large companies on the defensive. Despite these challenges, there is general agreement that outsourcing is here to stay, and its widespread use will continue and likely spread for competitive reasons. Although Asian countries led by China, India, and Taiwan have become the dominant outsourcing players, Eastern Europe has emerged as an alternative, and the former Soviet republics, Russia, Mexico, and Brazil are other potential outsourcing destinations.

Since outsourcing may not be the best option for some functions or products, a well-executed decision-making process based on a thorough analysis and proper selection of suitable partners and locations is critical. Factors that affect outsourcing decisions vary for systems, device, foundry, equipment, and subassembly/component suppliers.  Intel, Samsung, and other large IDMs with high-volume products do not outsource their chip manufacturing for their flagship products. On the other hand, a hybrid fab-lite or similar partial-outsourcing strategy has been adopted by many midsize and small suppliers. Some IDMs have even transitioned into a fabless model or are on the verge of doing so. The outsourcing of high-volume, noncritical modules, mature standard products or assemblies containing no core IP, and labor-intensive operations all reduce costs, making them common practices. For equipment suppliers, standard product designs or older and legacy products lend themselves well to outsourcing.

In a prudent approach, the nature of the project, the total cost of outsourcing (including management overhead and  hidden costs), its actual cost structure and all of its components, tax incentives, market access, low-cost local materials, project schedules, latency, quality, and the overall productivity must be carefully examined. Furthermore, a thorough economic analysis of no outsourcing versus offshoring and other various options (including different locations) must be performed over the duration of at least an entire business cycle. Any decision must also consider the human side up-front and properly deal with the workforce. Ideally, the product life-cycle should be managed and planned with outsourcing in mind right from the start. This approach would allow for deployment of a properly sized structure, avoiding the cost and pain of subsequent restructuring.

To manage risks, second-sourcing is essential, which may present a real dilemma for low-volume products. IP protection is another major issue that should be carefully examined. The transfer of IP and technology to contract manufacturers, combined with often weak and unenforceable laws and the confusing legal systems of some emerging offshore locations, have created a lot of concerns over outsourcing's long-term business implications. Many observers believe that this very issue has greatly contributed to the emergence of several strong companies over the past few decades which started as distributors or outsourced partners of OEMs. Pressed by their low margins and their desire for growth, some ODMs have begun to introduce their own brands, directly competing with their partners.

As some companies have learned, throwing the product over the wall to the outsourcing contractor, and assuming that heavy engagement in the early phases of the program is enough, can spell disaster. Risks include poor quality, delays, errors, inefficiencies, poor management of design changes, cost overruns, poor cost control, and inaccurate cost estimates and budgeting.  Other potential risks are poor communication, language barriers, cultural differences, currency fluctuations, time-zone differences, legal issues, and concerns over IP protection. Supply-chain issues—such as inaccurate forecasts, shortages, and the possibility of being put on allocation in upcycles—can also cause serious problems. Therefore, the active management of outsourcing and its potential risks, as well as its effective monitoring and tracking, are critical to the successful achievement of the targeted objectives and results.