The collaborative nature of conceptual modeling as an instrument for describing and changing the social world has been widely accepted. Despite this overall agreement there is little research on detailed and specific architectures for organizing the process of collaborative modeling and on tools that can support it.

Most studies take a certain procedure for granted: The modeling group consists of a facilitator (or modeling expert) and a number of domain experts that possess the required knowledge. The job of the facilitator is to “squeeze” out this knowledge and to cast it into models. The necessity of the facilitator is often motivated with the argument that the domain experts do not possess the required modeling capabilities and cannot sufficiently abstract from their concrete experiences. The other group members are therefore often reduced to a role of being passive information providers.

COMA improves this situation in the following ways:

• Group members are active and enjoy the model building process.
• Team members can work in a distributed setting partially of for the whole modeling process.
• The modeling process is documented completely.
• There is no need for redrawing paper-based models.
• The facilitator gets better and more structured input for modeling decisions.
• All group members contribute actively which increases model quality.
• Group members are satisfied and feel model ownership.
• This ensures stakeholder buy-in and therefore better project success.