Working across distance
The following tips for geographically-dispersed teams are excerpted from Professor Cramton's paper, "Information problems in dispersed teams," which appeared in the 1997 Academy of Management Best Paper Proceedings.

By Catherine Durnell Cramton

      Information technology has given new communications tools to teams, allowing collaboration to span great distances. For businesses with widespread operations, this is a welcome development. Through the use of tools such as electronic mail, videoconferencing, shared electronic databases and other groupware technologies, businesses hope to put the best person on the job while reducing time lost to travel and the need for relocation. However, some of my recent research has suggested that collaboration across geographical distance requires new team-building techniques and new team skills.

      I have analyzed the communication exchanged over a six-week period by members of thirteen geographically dispersed teams. This included 1,754 pieces of email, records of on-line "chats," the work products produced by the teams, and logs of their use of various communication tools such as email, telephone, fax and group decision-support systems. In every team, I observed destructive, self-reinforcing cycles of behavior in which information was conveyed poorly or lost, resulting in exaggerated or incorrect attributions about remote partners. I will describe four different types of information exchange problems that seemed to bedevil the geographically-dispersed teams: 1) failure to communicate information about context, 2) difficulty in communicating the salience of information, 3) unevenly distributed information, and 4) difficulty interpreting the meaning of silence.

      Failure to communicate contextual information. By definition, members of geographically-dispersed teams work from different locations. They may even be members of different organizations. There usually are important differences in their contexts and the constraints under which they operate. However, team members often failed to explain important aspects of their situation to their remote partners. Likewise, they seemed to find it difficult to create pictures in their minds of the contexts in which their distant partners worked. And when people did communicate some of their operating constraints to their remote partners, this information often was lost or forgotten. This created conflict, as remote partners failed to honor the deadlines of others, insisted on particular points seemingly without reason, or disappeared from the email traffic without warning.

      In light of this problem, I strongly suggest that geographically-dispersed teams take time at the outset to create a shared electronic "place" for the team, such as a team homepage or database. Team members in each location should carefully create there a picture of their situation: deadlines they face, criteria they must meet, times they will be away, the types of equipment or software they use and any constraints around access, hours they are most often available, travel time required, lead times required, alternative ways to reach each other, etc. While these constraints often are obvious and taken for granted in one's own location, they typically are invisible to remote partners. The homepage or database should be updated regularly.

      Differences in the salience of information. Teams members also ran into problems when they assumed that what was salient to them in a message would be salient to their remote partners. Sometimes, this problem arose when people sent email that addressed several topics. The most important topic in the eyes of Partner 1 may have been overlooked by Partner 2, who focused on some other aspect of the email. In addition, there was tendency to request feedback indirectly ("Hope to hear from you soon.") but to expect quick responses from each team member. They rarely came. I suggest that teams create shorthand that they can use to draw attention to the critical questions or points in the emails one is sending and to indicate questions for which a quick response is required from each member of the team. A glossary of these shorthand terms should appear in the team's homepage or database. If what you want is a quick response from a particular person or location, it is best ask directly and say how soon you must hear from the person.

      Unevenly distributed information. One of the most unsettling aspects of the team histories was the extent to which team members worked from wildly different sets of information without knowing it. This often put them on collision paths. Unevenly distributed information stemmed from a variety of types of errors and also from deliberate choices to send email to only a few members of the team. Errors included instances in which team members thought they sent email that did not go out, sent email that was lost in transit, and sent email to someone with an address similar to that of the intended recipient. The bucket was far leakier than team members realized. In some cases, erroneously addressed email was returned to the sender, but the sender never forwarded it to the original recipient; in other cases, damaging and erroneous conclusions had been drawn by the time the email was rerouted. Indeed, when team members finally discovered such problems, they often did not follow through by tracing in their minds any inaccurate assumptions they had drawn on the basis of information distribution errors.

      In other cases, people knew they had mailed only a part of the team but failed to understand how this practice affected the perspectives of team members who did not receive the mail. Sometimes, they later forgot that they had sent a message to only part of the group. There were serious consequences. Private exchanges of email distort perceptions of the volume of activity in a geographically-dispersed team. Those members who are receiving all the email perceive some members as active and others as inactive. Yet those who are not receiving all the mail perceive the energy level of the team as a whole to be low and the pace to be slow. They may further reduce their own participation or blame others for inactivity because they do not know of their efforts. The cycle can be self-reinforcing, as people send fewer messages to those perceived to be inactive, which gives the "inactive" members less to respond to and the impression that no one is acting.

      I recommend that teams establish a norm of copying all team members on email, while distinguishing between the primary recipient and those who were copied in order to keep them fully informed. In its database or homepage, the team should list its operating guidelines or norms, and these should include the practice of copying all team members. The team's glossary might include a term such as FYI (For your information) that members can attach to an email to establish that they do not expect a response from the person, but rather are sending it to keep the other informed. While this strategy helps solve one type of problem, it is not without costs. It can contribute to information overload and differences in what people find most salient.

      Interpreting the meaning of silence. One of the biggest challenges team members faced was interpreting the meaning of their partners' silence. Silence meant all of the following at one time or another: I agree. I strongly disagree. I am indifferent. I am out of town. I am having technical problems. I don't know how to address this sensitive issue. I am busy with other things. I did not notice your question. I did not realize that you wanted a response.

      One common problem was interpreting silence as consent when it indicated disagreement. People sometimes fell silent when they disagreed but did not want to confront the disagreement or could not find the right words to do so. If their remote partners assumed the silence meant consent, the two often collided down the road. I recommend that people refrain from jumping to conclusions about the meaning of a remote partner's silence and investigate the situation. Often the best way to do that is to use the telephone, which can help people convey and detect nuances that are lost in email.

      Information and interpretation. While flaws in the exchange of information such as those I've described here tended to set problems in motion in the teams, the problems usually were compounded by a tendency to judge remote partners quickly and harshly. This tendency may be stronger when partners are distant than when they are colocated. Therefore, it is a good idea for remote partners to meet face-to-face periodically or at the least to include some information on the team's homepage or in its database that helps personalize the members, such as photographs and a short list of interests. In addition, it is important to keep remote partners informed of one's activities and to give each other the benefit of the doubt in the face of sketchy or ambiguous information. Teams whose members did so and who set a gracious tone seemed to have a greater ability to solve the problems that inevitably would arise and proceed strengthened.

Updated 11/4/1998
Contact Information Copyright 1998 George Mason University