Beyond Related Experience: Upstream vs. Downstream Experience in Innovation Contest Platforms with Interdependent Problem Domains

Abstract

Problem definition: Innovation contest platforms are often organized around specific fields and host contests that span a variety of interdependent problem domains. Whereas contestants may benefit from related experience in contests whose problem domains share an interdependency with the focal problem domain, it is unclear whether the benefits of related experience arise symmetrically from upstream experience (i.e., experience in problem domains that provide input information to the focal problem domain) and downstream experience (i.e., experience in problem domains that use output information from the focal problem domain) or differ among them. Academic/practical relevance: Given that innovation contest platforms serve to effectively match contest problem requirements with contestants’ skills, it is important to understand how a contestant’s prior experience on a platform contributes to her problem-solving performance. Our research provides a more granular examination of the benefits of related experience than what has been examined in prior studies on individual learning or innovation contests. Methodology: We collected detailed archival data from TopCoder, a leading innovation contest platform that hosts contests across multiple interdependent software development problem domains, from its launch in 2001 to September 2013. Our data set comprises detailed participation histories of 821 contestants in 3,274 contests across eight interdependent problem domains involving 8,985 observations. Results: Whereas a contestant’s related experience on the innovation contest platform is more positively associated with her focal contest performance compared with unrelated experience, the benefits of related experience arise only from downstream experience. That is, there are no significant performance benefits of upstream experience. Furthermore, the performance benefits of downstream experience are greater when the contest duration is shorter, highlighting its role in enabling more efficient search and problem solving in innovation contest platforms with interdependent problem domains. Managerial implications: Contrary to the notion of “hyperspecialization,” our findings suggest that contestants can reap benefits from diversifying their experience into downstream problem domains on innovation contest platforms. Furthermore, innovation contest platforms could facilitate such targeted diversification of contestant experience by developing more granular metrics of contestant experience across problem domains. Our findings also have implications for resource allocation and job rotation decisions in software development projects within firms.

Effects of introducing an online community in a crowdsourcing contest platform

“…Third, we have developed a nuanced approach to examining the effects of experience and how it alters the effects of introducing an online community in a crowdsourcing contest platform. Prior research has demonstrated mixed effects from experience in crowdsourcing contests (Bockstedt et al, 2016; Liu et al, 2020; Menon et al, 2020). Thus, we followed other scholars (e.g., Bayus, 2013), who revealed the effects of multiple dimensions of experience (e.g., experience amount, experience diversity) in crowdsourcing.…”

Seeker Exemplars and Quantitative Ideation Outcomes in Crowdsourcing Contests

“…Our classification is based on the main research questions and objectives in the studies. Mihm and Schlapp (2019), Pollok et al (2019), and Wooten and Ulrich (2017); the second stream Brabham (2010) and Roth et al (2015); and the third stream Bockstedt et al (2016), Jiang and Wang (2020), Menon et al (2020), Seidel (2018), andSchemmann et al (2016). 5 The scan-shortlist-select process in our experiment also occurs in practice.…”

Designing contests for data science competitions: Number of stages and the prize structures

“…Regarding contest features, prize structure (Bimpikis et al., 2019; Korpeoglu et al., 2021; Nittala et al., 2022), information disclosure (Bimpikis et al., 2019), entry policies (Ales et al., 2021), and contest duration (Korpeoglu et al., 2021) have investigated. Other previous studies have utilized secondary data to investigate the impact of a particular contest feature such as contestants’ prior experience (Menon et al., 2020), contests’ problem specification (Jiang et al., 2021), and information disclosure (Wooten, 2022) on contest outcomes. Extending previous findings, our research leverages an incentive‐aligned laboratory experiment (e.g., Davis, 2015; Davis & Hyndman, 2018; Davis et al., 2021) to study the optimal contest design, which allows us to overcome several major challenges in empirically examining the effect of the contest design.…”