Given all the talk about the importance of creativity and innovation to the U.S. economy, it's amazing how little we know about the performance of creative people. How do you get the most out of people who are so often misunderstood as misfits and flakes? Starting with the basics: Do creative people work better in teams or all alone? And is it better to have creative teams competing with each other or all trying to help each other out?
Kevin J. Boudreau and Karim R. Lakhani, both of Harvard Business School, have come up with a very simple answer, based on an experiment using the same interface as TopCoder, a Web site where programmers compete to write software and algorithms solving real-world problems. The research suggests that creative people know very well how they work best, and will perform nearly twice as well if managers simply let them choose their own work environments.
The experiment involved more than 500 computer programmers, each of whom had a TopCoder ranking. Each programmer was paired with the programmer closest to their ability level. In each pair, one person was asked how they preferred to work: Alone, competing against 19 other people, or as part of a five-person team. The programmers were split almost evenly in their preferences. The second programmer in that pair was given the same work environment as the first person. The result is two equally-talented groups of programmers. The only salient difference between them is that one group is working in an environment of their choice, while the other group has had their work environment decided for them. Some of the participants were eligible for cash prizes, while others were not.
Participants then spent 10 days working on a problem for NASA. They were to write an algorithm that, given restrictions around size, weight, and versatility, would determine the optimal components of the space medical kit included in NASA missions. The winning algorithm is now in use on all NASA missions. Each programmer's algorithm was tested against 10,000 scenarios from NASA's database of 200,000 scenarios, and then assigned a numerical score based on how well it performed.
Teams vs. Individuals
The results underscore how important it is to let team players play on a team and to let loners alone:
- The programmers who got to choose how they worked produced algorithms that scored, on average, 83 percent higher than those who could not choose. That's about the same improvement as the researchers got by offering a $1,000 cash prize.
- Programmers who got to choose their work environment worked harder. The researchers measured this by counting the number of algorithms the programmer submitted. Programmers often submit more than one version of an algorithm, since there is no cost to doing so and having each version 'scored' provides valuable feedback.
- Those who got to choose their work environment worked longer. Without a cash prize, those who got to choose their work environment spent about seven more hours working that those who did not get to choose. If there was a cash prize, those who were in the work environment of their choice put in almost 11 more hours than those who had the choice made for them. The fact that folks who got to choose how they worked put in more hours accounts for almost all of the difference in their performance compared to that of their peers.
Kimberly Weisul is a freelance writer, editor and consultant. Follow her on twitter at www.twitter.com/weisul.