Here we illustrate the use of TeCFlow by analyzing four different social interaction scenarios:

Pattern 1:  Innovation in research and development: We analyze a globally active research and development community of a global management consulting firm,

Pattern 2:  Learning through online innovation dissemination: We analyze preparation and execution of a Web conference (“Webinar”).

Pattern 3:  Project management: We analyze communication in a distributed software development team.

Pattern 4:  Sales force: We analyze account management processes in a consulting practice focusing on one large Fortune 500 client.

Our sample data set consists of an e-mail archive of a virtual consulting practice with 200 members of a global consulting firm covering the time period from mid-2000 to early 2002. It is composed of the ego-networks of the practice leader and the practice coordinator (i.e. their e-mailboxes). Those e-mailboxes are taken as an approximation of the organizational memory of the consulting practice, as the practice leader and the coordinator were informed of all major events in the practice. The mailboxes were partitioned manually into mail folders by subject areas. Mail folders included one of eight service offerings, a folder for each project currently active, sales efforts, marketing activities, and the organization of two practice-wide seminars conducted over the Web (“Webinars”). The major advantage of this data set is that one of us was intimately involved in the analyzed work processes. The disadvantage is that the mailboxes of the practice coordinator and the practice leader do not include the direct one-to-one communication among the practice members bypassing the practice coordinator or the practice leader.

The TeCFlow tool is used in three steps:

(1)  Watch social interaction pattern movies to find dense clusters indicating potential emergence of COINs.

(2)  Look for peaks and troughs in the temporal evolution of group betweeness centrality and density to find changes in the collaboration patterns of groups.

(3)  Look at the contribution index to better understand the roles of the individuals in groups.

Combining steps (1) to (3) leads to new insights by giving an understanding of temporal evolution of group dynamics during the chosen time interval.

Innovation Pattern: Creation of a New Service Offering

This example illustrates the interaction pattern of an innovation community. It shows how to recognize the signal of a new innovation embedded into the general noise of communication. It includes 502 messages exchanged between 615 actors on general topics in the consulting practice over a period of 800 days. To spot innovation, messages have been searched and tagged for the first-time occurrence of the name of a new service offering with a unique name. The first time a message mentioning the name of the new service offering is sent to an actor, the actor is considered “infected” and remains colored red for the rest of the movie.

Figure 9. Creating an innovation (click here for QuickTime movie)

Figure 9 displays four movie snapshots visualizing four steps of the communication flow of the diffusion of the innovation. The cluster nurturing the new idea is clearly recognizable in all four snapshots, it takes about 400 days until the innovation is named for the first time and the cluster turns red. Afterwards the movie permits to track diffusion of the innovation from the original cluster to the remainder of the consulting practice.

Figure 10. GBC plot

Figure 10 illustrates evolution of group betweeness centrality (GBC) of the consulting practice. The first peak in GBC in figure 10 appears when the practice was officially started, GBC went up because of the broadcast announcements made by the leaders and coordinators to the entire practice. The second little spike is when the innovation team is at its peak, working feverishly and communicating with high frequency among themselves creating the new idea, which means that in relation to the entire social network centrality of the innovation subgroup is higher.

Figure 11. Contribution Index

The contribution index plot in figure 11 identifies the different roles in the practice. There are three individual roles of “practice leader”, innovation “creator”, and “practice coordinator” that stand out. There are also two clusters with similar contribution indices: the “marketing team” members, and the members of the “innovation team”. The “practice leader” is the most active actor, getting many more e-mails than he is sending. The “creator” of the new service offering comes second, demonstrating her immense efforts into the creation of this new service offering. The “coordinator” displays typical coordination behavior, sending more than he receives. The “marketing team” displays a “sales” pattern, sending more than they receive. The “innovation team” exhibits a “knowledge expert” attitude, receiving more than they send.

Main results of this visual analysis are:

*      Visual identification of the activities of an “innovation team”, coming up with an innovative new consulting service offering.

*     The identification of the time periods when the practice started and when the new “innovation team” was most active.

*    The central role of the “creator” in pulling together a team to develop this new service offering.

*      Occurrence of two clusters with holders of typical “marketing” and “knowledge expert” roles.

 Learning Pattern: Information Dissemination by Webinar

The next example illustrates how new concepts are taught to an online audience forming a Collaborative Learning Network (CLN) as introduced in section 6. The dataset for this example consists of e-mail messages on the subject of organizing a global Web-based seminar (“Webinar”); allocation of messages to the dataset was done manually. The archive includes 607 messages exchanged among 197 actors, covering a time period of about 190 days. The Webinar was prepared by a small team of self-selected members of the consulting practice over a multi-month period. One main speaker (the “practice leader”) then delivered the Webinar during one hour, assisted by his team members. The audience of the Webinar was spread out globally, and had the opportunity to ask questions to the speakers via e-mail during the talk. Questions that could not be covered during the talk were answered in the next few days. Because of overwhelming demand, the team decided to revise and rerun the Webinar a few weeks later. The team worked together on some minor changes, until the seminar was delivered again, this time coordinated by another main speaker, the “practice coordinator”.

The TeCFlow movie illustrates this switch between “innovation community” pattern and “learning community” pattern. In the preparation phase, the core team collaborated as a COIN, in the delivery phase, speakers and audience collaborated as a CLN.

Figure 12. Four screen shots of movie of Webinar communication flow (click here for QuickTime movie)

Figure 12 illustrates the changes in communication flow from preparing to conducting the Webinar. The picture in the top left of figure 12 shows the structure of the team preparing the Webinar. This group is operating as a COIN, with high density and low group betweeness centrality. The picture in the top right of figure 12 shows a screen shot of the communication pattern during the first time the Webinar was delivered. The “practice leader” (black dot) is sending and receiving information in a star structure. During and after this first run of the Webinar questions are asked to and answered mostly by the “practice leader.” The third picture at the lower left displays the team preparing a rerun of the Webinar, again working as a COIN and communicating with low group betweeness centrality. The final screen shot in the lower right of figure 12 displays the “practice coordinator” (blue dot) rerunning the Webinar, communicating with his audience and answering their questions in a star structure with high group betweeness centrality.

Figure 13. GBC of Webinar

Figure 13 illustrates changes in group betweeness centrality. The three phases in the organization of the Webinar can be recognized, with a decline in GBC (red line) for the organization of the first and second run of the Webinar, and a spike in GBC when the speaker delivers the event. In the preparation phase GBC is high because of the core/periphery structure of the core group leading a dialogue with potential speakers for the event.

Figure 14. Contribution Index of Webinar

Figure 14 shows that the “practice coordinator” is the most active, followed by the “practice leader”. The cluster of innovation team members working together to prepare the Webinar can also be identified.

Project Management Pattern ­ Communication in a Software Development Team

The next example analyzes a software project, where the consulting firm was acting as a general contractor to develop a bespoke software application on behalf of a client firm. The core team of the consulting company consisted of about 20 consultants, led by a project manager and a project partner. The client team consisted of the client project manager, a senior manager, and six subject matter experts.

The analysis described in this section (317 messages, 93 actors) focuses on the phase of the project right before “go live” and software handover to the client, During this period there were intensive negotiations between client project management and senior managers of the consulting practice. The following four snapshots illustrate the communication flow within the team in critical phases of this period. The top left of figure 15 shows the communication flow during the first phase when an addition to the original contract was negotiated by the “legal team”, while the technicians were working on implementing the technical system.

Figure 15 Four snapshots of project movie (click here for QuickTime movie)

In the top left window of figure 15 the “legal team” of the consulting firm forms a dense cluster, intensively discussing contractual details. The “consulting project leaders” as well as the client project leaders are in the center of the structure, communicating with everybody. The technical team members are forming another cluster at the top, mostly communicating among themselves. During the subsequent testing phase shown in the top right of figure 15, the “testing coordinator” has a centralized role coordinating the developers. The legal team is still tightly clustered, while the clients (green dots) are (too) peripheral.

During the next phase depicted in the bottom left of figure 15, legacy data is converted from the old to the new system. The “database administrator” is clearly recognizable in the center of the technical core team. The “legal team” is not active anymore, and client and consulting leaders collaborate somewhat more intensively. In the handover phase (bottom right) client and consulting leaders collaborate closely, while the technical team forms a separate cluster. The client technicians and the consulting technicians are not collaborating very much.

Figure 16 Group betweeness centrality of project team over 150 days

Figure 16 illustrates the main activities during this phase of the project. The first barely noticeable reduction in GBC is when testing starts. The next change in GBC is an activity increase during legacy data conversion while testing is still going on in parallel, leading to a decline in GBC. The final increase in group activity, leading to a reduction in GBC is during handover.

Figure 17 Contribution Index of different people in the e-banking project

Figure 17 illustrates the communication patterns of the different actors in the project. The consulting project leaders receive more messages than they send, and are also the most active senders. The programmers receive more than they send, while the administrative staff sends more messages that they receive. This is a fairly typical behavior for a commercial software project.

Findings from this communication analysis are:

• In the early phases of the project, senior leadership and technical staff on the consulting side are not well integrated. (During the project there was a major communication failure, leading to a delay and cost overrun. One can speculate that availability of such a “communications cockpit” might have increased awareness of communication problems, thus reducing chances of such breakdowns.)

• Critical roles such as “testing coordinator” and “database administrator” can be visually recognized. (During the project, junior project team members who assumed those critical roles were not appropriately recognized and rewarded. With a system such as TeCFlow such roles are quickly identified.)
• Some senior team members were not well integrated. (TeCFlow might have assisted in their integration.)
• The client was only peripherally involved in the handover phase. (Client staff was reluctant to assume responsibility for the system, resulting in project delays. TeCFlow could have pointed out this issue.)

As this example illustrates, TeCFlow could help in identifying critical issues in project communication, assisting project managers to better manage their projects.

Sales Force Pattern ­ Large Account Management in a Consulting Firm

Our final example illustrates the use of TeCFlow to analyze communication flow of a sales team. A new sales manager of an account management team of the consulting practice focusing on serving a single Fortune 500 client was taking over this responsibility from the previous sales manager who went into retirement. As his final assignment, the previous sales manager introduced the new sales manager to the client. Figure 18 displays four snapshots of the TeCFlow movie automatically generated from the e-mail archive (455 mails, 202 actors, 10 months).

Figure 18. “Sales” movie snapshots (click here for QuickTime movie)

The top left of figure 18 illustrates the old sales manager (blue dot) introducing the new sales manager (black dot) to the internal account management team. The top right of figure 18 shows a snapshot of the TeCFlow movie where the old sales manager is introducing the new sales manager to client executives. Consultants are red dots, clients are green dots. At this stage the old sales manager is still more and better connected to the client executives than the new sales manager, although new ties between the new sales manager and client executives are building up.

The bottom left of figure 18 illustrates coordination of a large proposal for the client where the new sales manager is not involved, as he is only peripheral to this cluster. The bottom right shows another proposal preparation where the new sales manager is in the center of the proposal team, while the old sales manager is in a more peripheral role. At the same time, a group of consultants and client executives is trying to arrange a social event, represented by the cluster of green dots at right, depicting this group of client executives. The new sales manager is too preoccupied with proposal preparation and misses this opportunity to connect with a new group of potential customers.

Figure 19 Group betweeness centrality of account management team

Figure 19 displays a history of changes in group betweeness centrality and density, identifying the major sales opportunities of the consulting sales team. Troughs in the group betweeness centrality curve indicate gatherings of a proposal team. As it turns out, the new sales manager only found out about some of these proposals through the TeCFlow analysis.

Figure 20 Contribution index of account management team

The contribution index plot (figure 20) shows that the most active consultant is a mid-level consultant working as a project manager at a client site. Fig. 20 also illustrates the somewhat passive communication behavior of the new sales manager. The old sales manager, although officially retiring from his function, is still communicating more actively. In addition, figure 20 also points out the two most active clients.

Findings of the TeCFlow analysis are:

• The old sales manager still held key client relationships even after the introduction of the new sales manager. The new sales manager should have been more active in leveraging the connections of the retiring manager in developing his relationships.
• The new sales manager missed a unique opportunity to connect with a group of new client executives, when they were trying to organize a social event.
• Coordination between different sales proposal teams was not optimal. The new sales manager missed multiple opportunities to support submission of proposals to the client.
• A mid-level consultant owned the best connections to the client. The sales manager could have made better use of these connections as a means of building new relationships.
• The most active customers were not known, and could therefore not adequately be taken care of.

A system such as TeCFlow might have helped to identify key clients and key proposal opportunities, thus potentially increasing the number of new opportunities for submitting proposals.

Why Time Matters in Social Networks

As has been shown in these four scenarios, the temporal visualization of social networks though movies of communication flows offers a novel visual way to discover different phases in the life cycle of an online community. It conveys insights that might be difficult to obtain by other means. The visual approach permits to find periods of low and high group betweeness centrality, and to identify potential periods of high productivity and information dissemination. It needs to be complemented by other contextual cues to obtain a full understanding of the activities, such as interviews with community members and a content analysis of the messages exchanged.

Users have found our movies intuitively useful to gain a quick overview of the dynamics of communication flow in groups. We are currently working on a more systematic study comparing analysis of longitudinal social networks by conventional means with our dynamic movie-based approach to get a more in-depth understanding of strength and weaknesses of our method.