Using the informal network matrix for project management teaming
James R. Kinney, P.E., Instructor, University of Alaska Anchorage
The successful project team must have the skills to perform the required work, and more importantly, must be able to function as a cohesive unit committed to achieving the project goals and objectives. Team members should have a high degree of trust for one another so that the individual's skills will be efficiently used. Without trust, work is continually scrutinized, and could be needlessly redone.
The purpose of this paper is to examine the role of informal network in project management organizations and ways to use this network to form cohesive, committed project teams. The paper builds on a case study (Krackhardt & Hanson, 1993) and its extended analysis (Ra & Fineman, 1998).
Informal Network Matrix
In any organization, understanding the flow of communication is key to successful management. An organization's informal network consists of communication lines, especially trust and advice between the various members of the organization. These lines form a network that can be completely different than the formal organizational structure. Exhibit 1 shows the formal organization chart for Krackhardt and Hanson's case study. Exhibits 2 and 3 are the advice and trust maps, respectively. The process of mapping the informal communication is relatively simple. First, separate questionnaires are distributed to all employees asking whom they go to for advice and whom they trust. They can select more than one peer, or no peers if that is the case. Second, the questionnaires are collected and the responses recorded. Third, the information is evaluated and interpreted; in simpler words, network analysis is performed.
In their case study, Krackhardt and Hanson used an example where CEO Leers appointed Harris as strategic taskforce leader. Harris seemed to be a good choice because of his work accomplishments, and he had the employee's respect. At first the task force seemed to be making headway, but after a few months it was apparent that no further progress had occurred. At that point, Leers mapped the advice and trust networks shown above and was able to ascertain that Harris's peers do not trust him, although his advice is frequently sought. Leers then appointed Benson, a person with considerable trust, as Harris's co-leader and the taskforce's progress resumed.
In their paper, Ra and Fineman demonstrated that each map (trust and advice) could be summarized in matrix form. The matrix uses a binary code, 1 or 0, to indicate if or not an employee named in vertical axis trusts an employee named in the horizontal axis. Exhibits 4 and 5 show Ra and Fineman's Advice Matrix and Trust Matrix, respectively for the aforementioned case study.
Matrix analysis allows not just an individual review of influence but also inter-and intradepartmental level examinations. To determine an interdepartment consult rate, divide the number of cells that have “ 1 ‘s” by the total number of possible links. For instance, the interdepartment trust consult rate between Software Applications and Data Control Systems is 1/16 = 6.3% in Exhibit 5. Similarly, the intradepartmental rate is calculated by dividing the number of cells in the department that have “1's” by the total number of possible links less the number of individuals within the department (self-consulting is not considered). So the Software Applications' intradepartment trust consult rate is 8/(16–4) = 66.7%. Exhibit 6 shows the organization's inter-and intradepartmental consult rates for trust. Advice consult rates by department can be also constructed in the same way.
In this paper, we use informal network matrix for selecting project team members. The first selection criterion is drawing team members into a project based on their specific skill sets. However, ideal project teams have members who mutually trust each other, especially a project leader trusted by all members. In order to demonstrate this process, we create a project led by Fleming from Field Design department who was appointed by Leers because of her special relation to a customer. This project needs four members, one from each department. Fleming wants to have team members whom she trusts and, at the same time, who trust her. Those members can be easily identified by looking at who have 1's in their respective row at the intersections with Fleming's column in Exhibit 5. They are Stewart in Software Application department; Swinney and Hoberman in Field Design; Muller, Daven, and Thomas in Integrated Communication; and Atkins and Kibler in Data Control. Exhibit 7 shows a short-version of the trust matrix comprising of Fleming and eight candidates. Exhibit 8 lists 12 (12=1*2*3*2) different team compositions. The more team members trust each other, the better teamwork will be expected.
Exhibit 1. Leer's Company Formal Organizational Chart
In Exhibit 8, matrices 3,4, and 10 show the most trusted team compositions having two out of six possible trust pairs for four members. Fleming can now select one of the three composition alternatives whose all members trust her as well as she trusts. In any team she selects, Fleming starts with the project with full knowledge of the informal relationship among team members. It would help the project leader accomplish many successes throughout project life.
Matrix analysis provides useful information to company managers. Network mapping identifies key employees who posses, individually and collectively, the appropriate mix of technical advice and trust links to accomplish project successfully. Matrix analysis allows not only an individual review of influence, but also inter-and intradepartmental level examinations. In addition, simulation and testing of organizational structure changes are allowable by mapping the informal network (Pleiman & Ra, 1999). If applied after the changes are implemented, network analysis can also determine how well an organization responds to such changes. Taking this one step further, periodic analyses can be used to rescore the organization and check progress on any “corrections” being attempted. This, in turn, can be used as a benchmark for instituting future changes.
Exhibit 2. Leer's Company Advice Map
Exhibit 3. Leer's Company Trust Map
Exhibit 4. Advice Matrix
However, there are some limitations associated with network analysis and its application. First, mapping of the informal network is only as accurate as the employee responses. If the completed questionnaires do not reflect reality then the informal network map and any interpretations based on the map are useless. Unfortunately this happens, especially when the employees think that management privy to the results may act unfavorably toward them if a perceived “wrong” response is given. Employees also have a tendency to feel they are more involved in the grapevine than they really are. An internal public relations campaign may be necessary to allay employee fears and support thoughtful responses. If handled appropriately, it should ensure that employees are honest in their answers. Employees must be assured that answers are kept confidential. Instead of portraying individual responses to questions, the company could adopt a departmental approach, as shown in Exhibit 6, thereby concealing individual responses but still being able to discuss the survey results. Another hurdle a manager may encounter is gaining sponsorship from upper-level management for a network analyst to ask employees personal questions about other employees and the organization (Stamps, 1997). Companies, like people, tend to be uncomfortable when their privacy is invaded. However, if a manager is successfully able to demonstrate the need and benefit of network analysis, upper-level management should be able to see through the initial awkwardness and sponsor the study. Finally, future analysis needs to identify the strength of each of the links on a measurable scale. For instance, Leers might seek advice from Blai five times a day and from Lang one time a day, but the study only shows that he seeks advice from both individuals, and not how often.
Exhibit 5. Trust Matrix
Krackhart & Hanson. (1993, July-August). Informal networks: The company behind the chart. Harvard Business Review, (104–111).
Pleiman & Ra. (1999). Restructuring the formal organization through informal network analysis. ESM 698 Report. University of Alaska Anchorage.
Ra & Fineman. (1998). Informal network matrixing: A new diagnostic tool for quantifying an organization's informal structure. ESM 684 Report. University of Alaska Anchorage.
Stamps, D. (1997). Off the chart. Training, (77–83).
Exhibit 6. Trust Consult Rates by Department
Exhibit 7. Trust Matrix of Project Team Candidates
Exhibit 8. Twelve Trust Matrices
Proceedings of the Project Management Institute Annual Seminars & Symposium
September 7–16, 2000 • Houston, Texas, USA