Team Collaboration in Innovation Contests

Abstract

In an innovation contest, an organizer elicits solutions to an innovation-related problem from a group of (external) solvers. Although solvers are capable of developing solutions individually and making individual submissions, they may collaborate as teams and make team submissions when encouraged by the organizer. Motivated by different policies adopted on various crowdsourcing platforms (e.g., Wazoku, Topcoder, and 99designs), we identify conditions under which the organizer can benefit from encouraging team submissions. We build a game-theoretic model of an innovation contest where solvers make either individual or team submissions. The quality of a submitted solution depends on the effort(s) of the solver(s) working on it and whether it is easy to decompose the problem to be tackled by different solvers submitting as a team. The quality is also subject to an output uncertainty, the level of which increases with the novelty of solutions the organizer seeks. We show that given a nondecomposable problem, an organizer benefits from team submissions when seeking high-novelty solutions (e.g., system design challenges at Wazoku) but not when seeking low-novelty solutions (e.g., logo design challenges at 99designs). We further show that when the organizer seeks (low- or high-novelty) solutions to a decomposable problem (e.g., software challenges at Topcoder), the organizer can only benefit from team submissions under certain conditions. We show the innovation-contest organizer should encourage team submissions when seeking high-novelty solutions to a nondecomposable problem or when seeking low-novelty solutions to a relatively difficult decomposable problem where solver effort is costly.

Keywords: Crowdsourcing, Platform, Team Submission, Tournament