In a study of 176 global semiconductor companies, researchers identified characteristics which predicted whether firms improved their technological performance by outsourcing the manufacturing of computer chips to outside companies.
"In the business press you often find individuals touting a one size fits all approach to outsourcing. Many articles describe outsourcing as a panacea for the high fixed costs and complexity of today's business world. However, this approach is misleading," said Michael Leiblein, co-author of the study and assistant professor of management and human resources at Ohio State University's Fisher College of Business.
"We identified the specific circumstances where outsourcing was successful for semiconductor firms- we believe that these circumstances may help managers in a variety of industries better understand when they can succeed with outsourcing and when it will be harmful."
Leiblein conducted the study with Jeffrey Reuer, associate professor of management and human resources at Ohio State, and Frederic Dalsace, from INSEAD, a French graduate business school. It will appear in an upcoming issue of the Strategic Management Journal.
According to Dun & Bradstreet's Barometer of Global Outsourcing, worldwide outsourcing expenditures were expected to top $1 trillion by 2000. Despite the rapid growth in outsourcing, there has been little rigorous empirical research that determines whether and when it is effective, Leiblein said.
In this study, the researchers studied cases where semiconductor firms from the around the world internalized or outsourced production of a variety of different semiconductor chips.
They wanted to examine how the relationship between the decision to outsource and the technological performance of the associated devices varied across different types of chips. Some researchers have suggested semiconductor firms would be more successful by focusing on their core competency - designing chips - and leaving the manufacturing to other companies, regardless of the characteristics of the product.
The results suggest a few basic principles which indicate whether a firm will benefit by outsource production, Leiblein said. "If you're going to outsource successfully, you need to know the critical characteristics of the end-product and production process, how to reliably measure your suppliers' performance in terms of these characteristics, and be able to predict how the suppliers' process will interact with other elements of your engineering and production system," Leiblein said.
For example, if a firm's product requires state of the art process technology or a great deal of customization in the manufacturing process, the intense coordination with the manufacturer is often both administratively expensive and technologically sub-optimal. "In cases like this, it would be difficult to use a supplier," Leiblein said.
In addition, firms need to be particularly careful about outsourcing when there is a great deal of uncertainty about future product demand. "If the environment changes after you've invested time qualifying a supplier, providing them with details of your product, and so forth, your supplier will likely have the upper hand in any renegotiation," Leiblein said.
The study found that when semiconductor companies outsourced their chip manufacturing under these circumstances the performance of their chips suffered. The basic principles were similar for firms operating in different regions across the globe.
"There are decision rules firms should follow that can help them decide whether or not to outsource," Leiblein said. "If you follow the model suggested in our study, you'll be able to make a better product."
While this study focused on semiconductor firms, Leiblein said the general results should apply to other technology-related industries. The bottom line, he said, is to align your outsourcing approach with the attributes of your product and your overall organizational strategy.