| Online innovation contest, a new online business model which can significantlyreduce R&D costs of enterprises and effectively take advantage of the intelligence of allparticipants, is emerging in recent years. This new model is called Crowdsourcingcontests in America and Witkey contests in China. As third parties, Crowdsourcing andWitkey websites are linkages between seekers and solvers, collecting the award fromcontests sponsors and distributing the money to winners. Although there have been lotsof studies on contest theories, very few of them are conducted to further understand thenew contest model and studies about the performance of online innovation contests arestill far from expected.In this study, several differences between online contests and traditional offlinecontests are identified. As a new form of contests, online contest can lower the cost andbring more participants from the whole world. Problem-solvers can take part in anyprojects without limitation. Besides, feedback can be given very often for onlinecontests. Then, several mathematical models are built and the number of participantsand the number of solutions are used to be the proxies of contests performance. Becauseof the different characters between online contests and traditional contests, onlineinnovation projects are classified into three types-ideation projects, expertise-basedprojects and ideation&expertise-based projects. Finally, how award design, along withother factors such as description length and contest duration, influences the number ofparticipants and the number of solutions is examined in this study.For this research, an empirical design is employed. All the data are collectedfrom one of the largest Witkey marketplaces in China-vikecn.com. On the platform ofvikecn.com, there are two kinds of prize structures. So a comparative study is exerted toexamine the different influences of two award structure on contests performance. A dataset of14317online contests is analyzed, with around4%multi-winner projects. Stata10is used to conduct the Pearson correlations, the White test and GLS analysis. Theresults show that more prizes a contest provides, more participants and more solutionsthere are. Likewise, longer contest duration also bring more contestants and more solutions. However, description length has no significant influence on most contestsexcept the ideation-based contests. At the same time, the influences of award structureare examined and the results illustrate that the structure of multi-awards is better thanthe winner-takes-all type for online innovation contests. The results about awardstructure are different from most studies on traditional contests, in which winner-takes-all structure is the best mechanism for award design. |
PDF Full Text Request