How often project team members had worked together was more strongly linked to team performance than the members' job experience in a study of Indian software development teams. What scientists call "team familiarity" was linked to fewer code defects and greater likelihood the project was completed within the labor estimates by themselves, or labor and schedule estimates combined.
Harvard Business School researchers Robert Huckman, Bradley Staats, and Daniel Upton found a source for all this data on more than 300 projects over 20 months in one company. Wipro Technologies is the third-largest software services firm in India, with 2006 revenues topping $3 billion. Wipro's internal databases allowed the researchers to calculate figures for data such as:
Higher familiarity related to fewer problems in every area except meeting schedule estimates. PM experience was linked to the degree of labor estimate accuracy, and engineer experience to a simpler yes/no measure of whether the labor target was met. (The authors said additional factors had significant correlations, but TeamResearch News uses a more conservative standard.)
PMs more familiar with team members might be better at assigning work based on member skills, the researchers suggest. People who know each other might be more willing to, and find it easier to, interact. Also, familiar engineers might be more willing "to put their reputations at risk by asking questions and sharing errors earlier in the process," they write. The authors suggest that managers might do well to consider familiarity, not just work experience, when putting together project teams.
Source: Huckman, R., B. Staats, and D. Upton (2009), "Team Familiarity, Role Experience, and Performance: Evidence from Indian Software Services," Management Science 55(1):85.