News Release

Key To Successful Software Development: Teamwork, Not High Tech, Study Shows

Peer-Reviewed Publication

Institute for Operations Research and the Management Sciences

LINTHICUM, MD, August 26--Software developers are more likely to succeed with team skill, managerial involvement, and a common level of team experience than with high tech tools, according to a study in the current edition of a journal published by the Institute for Operations Research and the Management Sciences (INFORMS).

The study also found that keeping information from clients at the vital early stage of software development damages performance.

The study is based on a 1994 survey of 66 software teams and clients from 15 organizations. The participating firms represented a range of industries, including insurance, financial services, high technology, and both heavy and light manufacturing.

The study examined the requirements definition stage of software development, the critical period when design teams interview clients and determine the software's eventual features. Although significant, the authors cautioned that their conclusions cannot necessarily be generalized to the entire development life cycle.

Avoiding Denver Airport Debacle

The study notes that there are numerous examples of highly visible software development failures. One of these was the one-year delay opening Denver's new international airport, at a cost of $1 million a day, due to a software problem in the automated baggage-handling system.

"These costly and conspicuous failures of software development projects point to a serious challenge for IS researchers," write the authors. "We must carefully examine the software development process to understand how to develop better systems."

Many Factors Studied

The study examined three types of behavioral factors and three technical factors to learn how the development team members would build a good working relationship and develop strong production processes. The authors also used these factors to assess processes that help the team coordinate with others. The study surveyed both team members and stakeholders - the eventual users of the software - to assess team performance.

The behavioral factors examined were (1) experience spread, i.e., a comparison of individual team members' length of experience developing similar software; (2) team skill; and (3) managerial involvement.

The technical factors studied were (1) structured methods, such as systems development life-cycles, information engineering, data flow diagramming, data modeling, structured programming, and object-oriented methods; (2) production technology, which is used to generate design decisions and subsequent products; and (3) coordination technology, which facilitates interaction among the planners.

Clear Implications for Management

One of the study's most important conclusions, the authors said, was that a very involved manager influences the types of processes associated with high performing teams in almost all instances. The study warned, however, against managers whose involvement becomes so intense and defensive that they fail to work well with those outside the team.

Technology's affect on team performance, in contrast, was less important. Structured methods were positively but not significantly related to production activities and had little affect on group processes. Coordination technology was unrelated to group processes or to performance.

Surprisingly, the authors observed, production technology was found unrelated to production processes and negatively related to relationship processes within the team.

"One of the most interesting conclusions of the study," wrote the authors, "is that the use of production technology actually interferes with the successful performance requirements determination of the tasks as represented by team members themselves."

The study confirmed previous IT studies reinforcing the importance of maintaining good relations upward in the organization. The study found that end users praised teams that practiced open communication and frowned on those that were secretive.

The research confirmed earlier theory about the importance of setting clear goals and milestones.

"When performing a task as complex as software development, team members must stay on track and achieve specific intermediate goals in order to increase their team's performance," write the authors. "The implications for management is clear: effective plans and procedures are critical."

The study, "Enabling Software Development Team Performance During Requirements Definition: A Behavioral Versus Technical Approach" was written by Patricia Guinan, Babson College in Wellesley, MA; Jay G. Cooprider, Bentley College in Waltham, MA; and Samer Faraj, University of Maryland, College Park. It appears in the current edition of Information Systems Research, a publication of INFORMS.

The Institute for Operations Research and the Management Sciences (INFORMS) is an international scientific society with 12,000 members, including Nobel Prize laureates, dedicated to applying scientific methods to help improve decision-making, management, and operations. Members of INFORMS work in business, government, and academia. They are represented in fields as diverse as airlines, health care, law enforcement, the military, the stock market, and telecommunications.

###



Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.