Innovation is a core driver of the U.S. economy. Organizations are constantly seeking to be more innovative with fewer resources. Advanced information and communication technologies that allow organizations to extract good ideas for products, services, and policies from the diverse suggestions of user communities represent one way to increase innovation without a concomitant increase in resource expenditures. But the outsourcing of idea generation to non-specialists increases the variance in the quality and applicability of the ideas that are generated and, concordantly, strains an organization's ability to give equal attention to all suggestions. The goal of this study is to explore how organizations might more effectively harness "the power of the crowd" to find and recognize the most promising ideas within a sea of contributions.
Recently, a good deal of research into team design has shown that the structural composition of a team can explain variations in the team's ability to generate good ideas. Conceptualizing a collective of crowd-soucing voters as an "emergent virtual team" allows us to explore a number of hypotheses about when the ideas that are voted to be "good" by emergent virtual teams with particular structures will, in fact, be the ideas upon which an organization should spend its limited resources of attention. In this two-year project, we collect and analyze ideas submitted and voted on by users in three community-based innovation forums, develop an algorithm for efficiently and effectively filtering good ideas from bad, and test the predictive power of the algorithm on other community-based innovation forums.
