Research projects in the engineering domain are often performed in partnership between academia and industry. Some kinds of funding presuppose a specific blend and participation of different partners in the projects. The initiator of a project, the prospective consortium leader, thus has to take on board a selection of organisations to meet the requirements from the funding programme or funding body. One usual way is to ask "old friends", partners you know from previous collaboration, but sometimes it is not sufficient to get a well-balanced consortium with sufficient competence and/or country representation. On the European level, there are many attempts to help proposers to build a consortium, often named "brokerage events", "proposer´s days" or the like. These often seem to encourage large consortia, both in principle and in practice. Large consortia can possibly help to be more complementary and well balanced, but there is no evidence that the quality of the project performance or result is better.The encouraged process can be described as a "snowball" method. New partners arrive more or less spontaneously, and in their turn bring more partners to the consortium. Often, groups of partners who already know one another tend to join a developing consortium cluster-wise, either by country or by research or competence area. From the funding programme or funding body, consortia are also encouraged to merge with one another, which gives the process of building a consortium another level of complexity. An advantage of this model can be that partners with previous experience join the consortium, group-wise.
An opposite approach is to be open, but to stay in control of the consortium-building process. This can be done by giving successive tasks and instructions to the interested partners, without any promises, and just successively incorporating new partners in the consortium. This strategy is a structured semi-open selection process for the consortium, paired with the development of the proposal.We state the hypothesis that the "standard" mechanisms might lead to large consortia, loss of control and possibly bad performance, and that a more restricted method can lead to sufficiently sized consortia, help the consortium leader to keep control, and possibly lead to better quality of performance. To learn more about this, we have interviewed a number of experienced project managers to find out how they perform consortium-building in practice. We have also been interested in learning more about how they examine new research partners before inviting them to a research cooperation proposal or project.