The Communication Risks on Project Success - A Review

Transcription

1 COMMUNICATION ON VIRTUAL AND CO-LOCATED SOFTWARE PROJECT TEAMS April H. Reed College of Business East Carolina University Linda V. Knight, School of Computing and Digital Media DePaul University ABSTRACT Communication plays an important role on project teams. This paper examines differences in communication issues on virtual as opposed to co-located software development project teams. A survey of over 150 IT practitioners was conducted to explore three specific communicationrelated project risk factors and their impact on the successful completion of software projects. Responses revealed that, while inadequate communication plagues all projects, the largest communication-related risk distinction between virtual and co-located teams lies in knowledge transfer. A plausible explanation is an inability to transfer certain types of implicit knowledge or expertise successfully without face-to-face communication or via electronic communication methods. KEYWORDS: Virtual Teams, Communication, Co-located Teams, Knowledge Transfer INTRODUCTION Several forces such as outsourcing, offshoring, availability of high-bandwidth communications and reduced business travel due to cost and security concerns have increased the reliance on virtual software development project teams[1]. Project teams are defined as groups of people working together to accomplish an overall goal while using electronic communication methods[5]. Virtual project teams involve team members who are not co-located, i.e. residing at different sites, cities, states or even countries. The prevalence of virtual teams raises questions about differences in performance on virtual versus co-located teams. One area of potential difference is the degree of impact caused by various risk factors. Project risk is linked to project failure or success because risk factors on projects can evolve into issues [2]. Those issues can result in problems ranging from loss of dollars, loss of work effort - 1 -

2 or lost opportunities as well as complete project failure. Identification and knowledge of project risk factors has been cited as a method of decreasing the severity and the impact of risk [2]. One risk factor that has been identified as a top risk factor is communication [9]. This paper will explore the differences in some specific communication-related risks on virtual versus co-located software project teams. The main purpose will be to answer the following questions: 1) Are there significant differences between the two types of teams and; 2) what are the implications of any such differences. Previous research has focused on the complexities of communication on teams. Three specific project risk factors relating to communication were distilled from this past research, along with a combination of focus groups and pilot studies. Note that the focus group and pilot studies are described later in the methodology section. One project risk factor that emerged from this background work, Lack of or inadequate communication, focuses on a low level of communication frequency or communication at the wrong level of detail for the audience. For example, a project dealing with familiar processes and well known work might require a low level of communication frequency since team members are experienced. On the other hand, innovative or technically challenging projects might need a high level of communication frequency to deal with many unknowns. Jones, et al. (2004) indicate effective communication is the most critical component of teamwork [5]. Wallace et al. (2004) in their research on outsourced projects, which is a type of virtual project, indicated team risk may be due to greater challenges in team communication and coordination, especially when at least two organizations were involved [10]. The second risk factor studied here, Technical connectivity issues hinder communication, focuses on technical issues with communication tools used by virtual teams. Since those teams are dependent on electronic communication, any down time could effectively isolate members of the team. This risk factor was identified from face-to-face interviews and a focus group, as detailed in the methodology section. The third risk factor studied here, Insufficient knowledge transfer, refers to a more complex aspect of communication. Knowledge transfer is defined as the unidirectional exchange of knowledge. These transfers generally have a clear objective and are geared toward a specific recipient[6]. However, knowledge transfer is much more complex than the definition indicates. There are two distinct methods of knowledge transfer, explicit and implicit. Explicit transfer is more formal and uses methods such as documentation, training and interviews. Implicit transfer is very informal and involves methods such as storytelling, mentoring or coaching, and communities of practice[4]. Wallace (1999) identified this risk as an important part of team risk dimensions in her research on traditional software project risk [5]. METHODOLOGY - 2 -

3 A survey of over 150 experienced IT industry practitioners was conducted to identify the degree of impact specific risk factors had on the successful completion of software development projects. Fifty-five risk factors, three of them centering on communication, were identified through the literature, face-to-face interviews, and a focus group of practitioners. Once identified, the risk factors became part of a survey questionnaire, where participants were asked to focus on a recent specific virtual software project on which to base their answers. Those who had not participated in a recent virtual software project were asked to answer based on a recent co-located software project. Survey participants were asked to respond to each risk factor using a three-point Likert scale to indicate the degree of impact each risk factor had on the successful completion of their specific project. In the three-point Likert scale the possible selections were major impact, minor impact or no impact/ did not occur. Overall, about two-thirds of the projects on which the survey was based were virtual software development projects with the remaining third being co-located software projects. A majority of the projects were new development types (48%) with the remainder being software upgrades and package upgrades. The duration of a majority of the projects (61%) was under 2 years. About a third of the project teams each (31% and 30%) contained more than 20 people or 6 to 10 people respectively. Project costs varied widely with a third (33%) of the projects in the over $1 million range and approximately another third (30%) in the $100,000 to $500,000 range. Finally, participants were practitioners such as IT project managers, systems analysts and IT project leaders, but not end users who came from a wide variety of industries. RESULTS AND DISCUSSION The communication related hypothesis is as follows: Risk factors related to inadequate communication will be significantly higher on virtual software projects than on co-located software projects [8]. Three of the fifty-five risk factors in the survey were communication related and are listed in Table 1 below. Also included in the Table 1 are the participant response percentages which were collected for each of these three communication-related risk factors. VIRTUAL PROJECTS CO-LOCATED PROJECTS Risk Factors No Impact or Did Not Occur Minor Impact Major Impact No Impact or Did Not Occur Minor Impact Major Impact p-value Lack of or inadequate communication Technical connectivity issues hinder communication Insufficient knowledge transfer * Table 1: Communication-related Risk Factor Survey Response Percentages - 3 -

4 The participant response percentages from the first communication-related risk factor indicated the risk of Lack of or inadequate communication was thought to have an impact on the successful completion of both virtual and co-located projects. Almost half (48%) of virtual project participants and 40% of co-located project participants indicated this type of risk had a major impact on their project. About a third (35%) of virtual project participants and 36% of colocated project participants felt this risk caused a minor impact. Additionally, this risk factor scored the highest on major impact of the fifty-five risk factors in the survey. These results are understandable and are in line with other literature which indicates communication is an important risk on a project [7]. What is surprising is the insignificant difference indicated by a p- value greater than 0.05 (0.6267) between virtual and co-located projects for this first communication-related risk factor. The lack of significant differences for this risk factor may be due to several things. First, it was thought that the physical distances between project team members could understandably create communication lapses. Instead, it appears lack of communication or inadequate communication is a major problem regardless of the proximity of the team members. However, the major impact was higher on virtual projects than co-located projects. This may occur if virtual projects are managed using the same methods as are used on co-located projects where face-to-face meetings are often used as a more common form of communication. The lack of face-to-face communication on virtual projects may be unsatisfactory to the team members who are used to that method. It is also possible that some team members will never develop a comfort level with dependence on electronic communication to run a project. A new model of communication may be needed for virtual teams which addresses the richness of the content and a variety of communication methods depending on the frequency and goal of the communication. The participant response percentages to the second communication-related risk factor indicated Technical connectivity issues hindering communication did not have a major impact on the surveyed projects. Only 15% of virtual project participants felt the impact of connectivity issues was major, while 53% felt the impact was minor. On co-located projects only 17% of participants felt the impact was major, while 49% felt the impact was minor. This spread of percentages may be an indication that connectivity issues have stabilized over time and now are rarely a cause of major impact on a project. Both the availability of cheap telecommunication bandwidth and the expertise to maintain its stability may be the cause of the lower level of impact. It is possible that technical connectivity issues still occur, but the expertise to correct them is prevalent and the availability of backup systems exists. It is not surprising that the differences in participant responses between virtual and co-located projects for this risk factor are not significant with a p-value greater than 0.05 (0.8783). This lack of significance is probably also due to the widespread availability of cheap telecommunications bandwidth. Technology problems no longer appear to be a major factor in hindering project communication. The participant response percentages to the third communication-related risk factor, Insufficient knowledge transfer, widely varied. Figure 1 below shows the participant response percentages for each impact level and for both virtual and co-located or traditional projects

5 Figure 1: Impact on Project of Insufficient Knowledge Transfer This risk factor had more of a minor impact than a major impact according to the responses of all participants. However; the impact of insufficient knowledge transfer was significantly greater on virtual software projects than co-located software projects with a p-value well below 0.05 (0.0023). Almost twice as many (62%) co-located project participants felt this risk factor either did not impact the project at all or did not occur on their projects. Only 32% of virtual project participants concurred. Unlike the other two communication risk factors, where no significant differences were found between virtual and co-located project groups, the differences between the two groups for this risk factor are significant, with a p-value of These results indicate knowledge transfer is more of an issue on virtual projects than on colocated projects. To understand why this may be the case we need added research comparing knowledge transfer on virtual and co-located projects. Knowledge transfer often centers on the sharing of expertise gained from years of experience[3]. It is likely some of this information is not documented, but exists in the minds of seasoned employees. A common formal method of transferring knowledge is through cross-training or shadowing of an experienced employee through the use of face-to-face communication and meetings. In addition, knowledge sharing on co-located projects that is implicit undoubtedly takes place informally, through water cooler or over-the-cubicle remarks. This type of casual knowledge sharing does not really have a clear virtual equivalent. Sharing of undocumented knowledge and face-to-face exchange of information can be difficult to accomplish in a virtual environment. A possible solution is to convert more information into explicit knowledge and to emphasize documentation, which can then be exchanged electronically. However, this does not resolve the exchange of implicit knowledge, which will always exist and may be just as important as the explicit knowledge. CONCLUSION In conclusion, a comparison of three communication-related risk factors has revealed some significant differences between communication on virtual and co-located software projects. In the cases of inadequate communication and technological problems hindering communication, the location of the employees appeared to have little influence on the occurrence of the risk. In - 5 -

6 other words, being a virtual project did not prevent the risk from occurring. However, knowledge transfer was found to have a significantly stronger negative impact on virtual software projects than on co-located software projects. A plausible explanation is an inability to transfer certain types of implicit knowledge or expertise successfully without face-to-face communication or via electronic communication methods. This is exacerbated by the failure of the virtual environment to provide substantial opportunities to provide for the casual exchange of knowledge. More research is needed specifically on differences in knowledge transfer between virtual and co-located project teams, and in ways to enhance implicit knowledge sharing on virtual projects. [1]. Aspray, W., et al., Globalization and Offshoring of Software: A Report of the ACM Job Migration Task Force, W. Aspray, F. Mayadas, and M.Y. Vardi, Editors. 2006, Association for Computing Machinery (ACM). p [2]. Boehm, B.W., Software Risk Management: Principles and Practices. IEEE Software, 1991(January 1991): p [3]. Choi, C.J. and S.W. Kim, Women and Globalization: Ethical Dimensions of Knowledge Transfer in Global Organizations. Journal of Business Ethics, : p [4]. DeLong, D., Lost Knowledge: Confronting the Threat of an Aging Workforce. 2004: Oxford University Press [5]. Jones, R., R. Oyung, and L. Pace, Working Virtually: Challenges of Virtual Teams. 2005: Idea Group Inc. [6]. King, W.R., Knowledge Transfer. Encyclopedia of Knowledge Management, ed. D. Schwartz. 2006: Idea Group Publishing [7]. Pare, G. and L. Dube. Virtual Teams: An Exploratory Study of Key Challenges and Strategies. in Proceeding of the 20th International Conference on Information Systems Charlottee, North Carolina: Association for Information Systems. [8] Reed, A., Differences in Critical Risk Factors on Virtual Versus Traditional Software Projects, 2008, Depaul University, College of Computer Science, Telecommunications and Information Systems, Chicago, IL [9]. Wallace, L., The Development of an Instrument to Measure Software Project Risk, in Computer Information Systems. 1999, Georgia State University College of Business Administration: Atlanta, GA. p [10]. Wallace, L. and M. Keil, Software Project Risks and Their Effect on Outcomes. Communications of the ACM, (4): p