The software outsourcing has already become a trend. Today, the progress of software technique is getting more and more rapid, and the computerized requests have increased too much more. Each trade, confronts the demands of software, has gradually thought about going on with outsourcing. This research is with such a starting point, to study the outsourcing projects, in order to find the way to succeed in outsourcing practices. The steel industry is a typical manufacturing industry of Taiwan, which the computerized degree has not been popularized like others. This research is to regard steel industry of Taiwan as the target of studying. Choosing the best practice of this environment, and through the implementation experiences of best practice, it comes to study what factors influence whether the software outsourcing project succeeds or not. This research combines the Institutional Theory with work standards of the software outsourcing, as the foundation of the theory. And then cooperate with the four stages, planning, assessment, control, confirming, to marshale the external institutions and formal internal institutions of the software outsourcing in theory; Equally, according to the current practices of best practice, separating into the same four stages, get the formal and informal internal institutions and external institutions, which implement in the best practice. Finally, verify the theory and the practice, to analyze what¡¦s the difference. By the experiences and implementations of best practice, we can verify the factors that influence the software outsourcing. Besides to clearly define the internal and external formal institutions, we still need the informal institutions as an auxiliary. It will be a great benefit to the software outsourcing. Of course, the company must define the institutions as could as possible to the limitation that the formal institutions can be. And try to make them be worked. But many informal institutions play the key role behind. Without the informal institutions, the software outsourcing projects might move towards failing too. If the formal and informal instituteions can complement each other, the failing rate of a software outsourcing project will be minimized.

Software development outsourcing is a very common way to gain source for information system among government and private business. Although there are lots of advantages of software development outsourcing, if one do not really understanding outsourcing correctly, not only expected results cannot be met, there will be additional problems. From the past research data and real cases, there are lots of problems shown. There are always lots of participants involved in product development, thus conflicts are very closely associated with participants. If conflicts do not get solved immediately or root caused had not been found right away, conflicts always results in large damages to developments. For example, there are hidden conflicts with in ambiguous document format and request. When mistakes stack up and do not get solved right away, they will become potential conflicts. However, most members do not pay too much attention to the reasons how conflict occurred or to ways of solving but pay more attention on teamwork and professionalism. There were lots of research done in the past regarding the conflicts between customers and suppliers of software development outsoursing; most of them are limited in reseach conflicts of orgination and individuals. Most research methods are market research or individual interviews there is no in depth research regarding the conflict events with long period of data collection and analysis; also tracking the major factors for the root causes. This research method involoved two real cases. We observe, record, and collect data from the real events and interviews participants. We investigate the causes of conflicts between customers and outsourcing suppliers by framing theory. Then we analyzed the cognition of customers and outsourcing suppliers toward conflicts during the research period. Then, this research result came from using six main conflict frames work as conflict types : identity, characterization, fact, power, loose versus gains, process. After the data was collected, we use consistent comparison as a foundation. The purpose is to cross exam the data and found the root causes before and after conflict resolution. In software industry, when people encounter conflicts or found problems of conflict, this research helps them to understand the root cause faster and it provides great resource for project managers. By using real cases records and results as examples to find the root cause or potential problems; people could come to neutral agreement by using conflict management to satisfy both parties.