The Idea:

Individual incentives based on impact yield more and better ideas in group settings.

The Research

Olivier Toubia used an online discussion tree to devise an impact-based measure of how one idea in a group setting affects the generation of other ideas. That measure then served as the basis of an incentive system to reward individuals for stimulating ideas among other group members. An online test of the measure, with three groups of 26 members each, showed its value compared to two more common incentive methods.

The test asked the groups to generate ideas on the same problem — in this case, how to improve the United Nations Security Council. In the first group, everyone received the same flat fee. In the second group, your reward depended on how many ideas you contributed. In the third group, your reward depended on how many ideas other group members added to your ideas in the discussion tree.

The results showed that the third group generated more ideas overall, more different streams of ideas and ideas that were more fully articulated.


Practical Applications

Creative marketing professionals

You can use an impact-based incentive method to generate more and better ideas in every phase of your marketing program. Because we all increasingly exchange ideas by e-mail, this method is easy to implement and easy for everyone to use. The research method includes ways to analyze the data you get, beyond deciding on rewards, so you can adjust the method as you proceed to arrive at the best results.

Leaders of group strategy or design sessions

The study’s test problem — improving the U.N. Security Council — shows clearly that the impact-based incentive method can go far beyond marketing to all aspects of strategy and design. Its online nature allows group members to join in from all corners of the globe, over a day, a week, a month or any period of time you choose. And it offers a way to dramatically increase the number of people you can bring into strategy or design sessions, with low cost and high impact.


© Marketing Science, 2006
Volume: 25 | Issue: 5 | Pages: 411-25

Publication type: Journal article

View abstract/citation