Asymmetrical

Recently, someone in a discussion commented about a workshop we participated in and suggested “it wasn’t well prepared in advance.” Indeed, the one day workshop practically wasted the time of some fifty people coming from all over Europe and Africa, spending valuable time on basic coordination activities and introductions. Post meeting, email discussion died off within three months, yielding no result. The email mode of communication was simply too restricted for the type of coordination that should have happened in advance and during the workshop.

Guess we all heard it before: “organizer(s) should prepare in advance for a collaborative event;” “a small group of people should do the homework before including a large group of people.” It holds for successful open-source projects as well: “if one wants to start a new project, one has to do most of the work before inviting contributors to join.” Even Kevin Costner had to build it first.

The amount of preparatory work is not symmetrical for a given collaborative interaction. The organizer does most of this work, and participants are expected to process it and give feedback, which is much smaller in volume of information per person. This is very much true on the edge of network. I usually describe it as small group inside engaging large group outside. In other words, the edge of network is about asymmetrical collaboration. Not just because success requires preparation; the communication channel is very much restricted across the edge, and one wants to ensure the focus most important aspects, whatever the real reason was to engage those outside people. Conversely, the small group inside communicates face to face and can process much larger volumes of material and prepare it in the appropriate form for others.

In terms of pure bandwidth spent, in most collaborative communities we see 70%-90% volume going out (usually in the form of downloads of the initial material), and only 10%-30% coming in from the outsiders. Even in an email discussion, a moderator will usually send out more messages compared to other participants. This asymmetry has some interesting design implications for extending the platform into mobile context.