Uraveling start-up processes with the help of sequence analyses

Our sequence analyses of the PSED2 database, the largest available dataset on venture creation processes, demonstrate two points. First, they show when and how to use the numerous variants of this method. To this end, we develop a decision tree that makes the analytical choices to be taken explicit. Since researchers can often not know ex ante which way of running sequence analyses delivers the most insightful results, we suggest to use this decision tree and: `in case of doubt, do it both ways!`. Second, our analyses also highlight the usefulness of sequence techniques for studying venture creation processes. Contrary to previous start-up analyses, we succeed in identifying 16 distinct ways in which entrepreneurs set up new ventures. These findings suggest that previous studies fail to recognize systematic venture creation patterns due to the use of traditional statistical techniques which do not make it possible to treat one sequence of events as one single case.