The impact of contest attributes on solvers' participation and performance in crowdsourcing contests

  • Fabian Kutschera (Student)

Student thesis: Master's Thesis

Abstract

Crowdsourcing contests as a means to leverage the knowledge of external individuals to solve firms’ innovation problems are growing in popularity. Little research has been de-voted to the question how such contests should be ideally designed in order to achieve a high level of participation and high quality solutions. By analysing a data set of 913 soft-ware engineering contests from the platform TopCoder, this study investigates the impli-cations of three particular contest elements: the number of participants, the monetary award and the problem description. The results confirm the theoretic model, stating that a large number of participants is preferable for a seeker (firm) and outweighs the arising disadvantage of lowered efforts of solvers due to a greater competition. This theoretic discussion was enriched by the finding of a curvilinear relationship between the number of solvers and the contest’s performance, wherefore adding contestants is only beneficial until a certain level. Furthermore, results show that a high monetary award captures more participants but leads to a lower contest performance, which is contrary to prior research. Regarding the problem description, it was shown that a short problem description in-creases participation and improves the contest performance whereas a problem descrip-tion, which includes URL links to other homepages, reduces participation but improves a contest’s best solution. By understanding the outcomes of this study, firms better under-stand how to design crowdsourcing contests.
Date of Award18 Jul 2016
Original languageEnglish
Awarding Institution
  • Universidade Católica Portuguesa
SupervisorCláudia Isabel de Sousa Costa (Supervisor)

Keywords

  • Crowdsourcing
  • Open innovation
  • Innovation
  • Contest elements

Designation

  • Mestrado em Gestão: Programa Internacional

Cite this

'