Unleashing the Crowd

We describe our field research in which we compare the rated innovativeness of crowds who were exposed to the Idea-Sharing process against “unmindcuffed” crowds. To just collect the data, we had to embark on a painstaking five-year effort of data collection and analysis because all third-party vendors were using the Idea-Sharing process. We describe details about how we ran the 20 crowdsourcing events in the field with partner organizations, making changes to third-party software to create the condition of unmindcuffing the crowd. We randomly assigned events to either the Idea-Sharing or the Unmindcuffed condition. We had executives rate the ideas for their novelty, implementability, and competitive advantage. We found that the unmindcuffed crowds produce more innovative solutions than the more traditionally used Idea-Sharing process!

This first practice describes how the crowds participated. Our findings indicate that individual participants, on the average, offer fewer than two posts! Their posts are short knowledge fragments (such as only a few sentences). They are not engaged in extensive back-and-forth questioning. They offer minimal social support for others’ posts. It is almost as though the posts serve as knowledge batons in a knowledge relay race. The participants will not be the same throughout the race. As in a relay, the knowledge posts act as knowledge batons passed from one leg of the race to another, where the next participant is whomever is interested in getting the baton with no direct communication between the runners themselves. Each participant just takes the baton and keeps the innovation process moving forward.

We coded the knowledge traces contributed by the crowds. We were looking for the presence of any of four different types of knowledge shared about the problem description and solutions: facts, examples, paradoxical objectives, and ideas for solving the problem. We found that innovative ideas were NOT preceded by a larger variety of ideas! Instead, innovative ideas were preceded by the crowd posting a greater variety of different knowledge TYPES. Thus, it is not simply diversity of opinions that matter, it is the diversity in how each member frames their knowledge when they are sharing it during the crowdsourcing.

Innovative ideas were more likely to be preceded by creative associations, than any other type of knowledge. Creative associations are personal experiences a participant might have had with a similar problem in a different context, or conflicting objectives a participant believes must be solved by any proposed solution. Creative associations have a twofold inspirational value. First, they inspire others to post their own creative associations, thus amplifying the number of creative associations available for anyone reading the posts. Second, they inspire the creation of innovative ideas by helping participants make connections in their own minds that spur creative discovery. This effect only works when the creative associations are present in the most recent five posts prior to the innovative idea; posts further back are ignored.

In this chapter, we examine the participants’ behaviors during a crowdsourcing event undertaken by a managed health-care insurance company. The executives conducted an internal crowdsourcing event to solicit from the claims professionals innovative ways to service customers. We found that those posts initially offering a specific need coupled with a specific solution garnered the most attention by participants, but did not necessarily lead to innovative solutions. What led to innovative solutions were participants peeling away the need, then reconstructing the need into a creative association only tangentially related to the initial need. Thus, innovation in crowds does not emerge from simply listing requirements for the solution, but rather, from needs being creatively reconstructed to inspire others to think of new solutions.

In traditional Idea-Sharing crowdsourcing, crowds are told what role to play: they are to be either Ideators (offering Ideas) or Idea Commenters and Refiners (offering comments to help others’ idea be refined). People who don’t have ideas, but just have thoughts, personal experiences, and assumptions about the nature of the problem are left OUT, without a voice. But such people are the bedrock of innovation in crowdsourcing since they provide the creative associations. In unmindcuffed crowds, some participants just offer facts while others just offer creative associations, and others just offer ideas. This allows everyone to have a voice.

Since we are interested in progressing research, we present a scholarly version of our theory of collective production of innovation in which innovating crowds consist of some participants willing to use their scant two posts to disaggregate their knowledge into creative associations of knowledge batons and others willing to take those knowledge batons and co-mingle them to stimulate creative discoveries. The disaggregation occurs as people break down their causal models, their coherent perspectives, their proposals of need-solution pairs into factual assumptions, short statements of ideas, and creative associations. Since crowds spend so little time contributing to the wicked problem, the more effective the crowd can be at eliciting each other’s disaggregated knowledge in a way that stimulates creative thought in a virtuous cycle, the more likely that the crowd will successfully produce an innovative solution. The implications for a new direction for research on innovation and new organizational forms are discussed.

How should the technology platform look like in order to unmindcuff the crowd in a way that collective production occurs? Since our initial search for technology platforms did not turn up any platforms that unmindcuffed the crowd, we describe the crowdsourcing platform interface which should be in place, which can make the difference between a less productive Idea-Sharing crowd and a crowd with the capability of collectively producing innovative ideas. We frame our guidance in terms of design principles, that is, guidelines for ensuring that the platform will promote collective production.

The ability for anyone anywhere to connect, and the access to data that this provides gives people the power of knowledge: to offer what they know about any wicked problem they are interested in, and join whichever flash crowd that suits their fancy. This has huge implications for research and management practice about organizations of the future. Managers will increasingly be expected to have a new skill: solving problems not with teams, but with crowds of anonymous, unrelated actors contributing only two posts. Researchers will need to reconsider the role of social interaction in how organizations function since social interaction is no longer as essential as it has been in the past. Strategists will need to think through which questions are best solved by different types of crowds (e.g., internal-only, external-only, hybrids, etc.) based not only on intellectual property issues, but on the ability of the organization to support the crowd’s capacity to innovate. Those organizations better able to build such a capacity in flash crowds will be more agile and effective in solving the wicked problems we face today and into the future.