Trajectories and Transitions of Entrepreneurial Groups

This sub-project aims to explain how the initial constellation and relations of entrepreneurial groups endure, change and dissolve over time. In an ambitious methodological design, four cohorts of entrepreneurial groups shall be tracked over time. Following the example of early studies on inter-locking directorates and industry networks, the proposed research project manually builds a data set from publicly available data leveraging computational techniques and the advancement of online-supported crowdsourcing

The German Commercial registry forms the universe from which we drew our sample. We collected all registry notification in the years 1997, 2002, 2007 and 2012 (N ~400.000) and drew a random sample of 2.500 cases for each cohort. For these cases we extracted information from the registry and commercial business data bases (such as ORBIS or NORTHDATA). On these grounds we can describe the number of people engaged, their roles in active management and ownership stake and some basic demographic information (e.g. age, gender). Eventually we can describe typical patterns in group composition. From here, we collect annual information of changes in the distribution of roles, ownership stake and the size of the group. This allows us to describe a sequence of the group’s trajectory and identify typical transitional phases.

For a subsample we apply a definition of entrepreneurial groups that does not equate the group with a single organization, but allows for a group to be involved in multiple businesses. The operationalization requires a complex search procedure for co-ownership and co-management of group members. For this subsample we extract additional information on the careers of involved group (e.g. from Linked-In and Xing) and on their social relationships (e.g. from websites and newsarticles). To contact the later information we set up a crowd science project: We established the platform “Datenspuren” on which we offer a 45 min online training on process generated data including a search exercise. Participating students could continue to perform the search exercise, collect points and win prizes. Through their performance students would help to crowdsource our data. The platform is available at www.datenkunde.org. We thank all participating students for their efforts and congratulate our winners!

Back to Top