Rank less, bond more: rethinking performance feedback
A new study from Cornell SC Johnson College of Business reveals that sharing performance information among employees can foster cooperation, but only if the circumstances are right.
The research, conducted by Xinyu Zhang, assistant professor at the Samuel Curtis Johnson Graduate School of Management, part of SC Johnson College, adds a new dimension to the understanding of workplace motivation, showing that performance feedback isn’t just about driving individual effort; it can also shape team dynamics. The paper, “The Effects of Relative Performance Information on Subsequent Cooperation,” appears in the latest issue of Contemporary Accounting Research.
“My study reveals that when workers receive feedback indicating they performed similarly to their peers on a difficult task, they are more likely to collaborate afterward,” said Zhang. “However, if the feedback includes rankings or highlights performance differences, it can reduce the sense of connection and hinder teamwork.”
The findings come from two experiments designed to explore how relative performance information (RPI) or data showing how individuals perform compared to others affects social bonding and cooperation.
In the first experiment, 192 undergraduate students at the University of Texas at Austin completed a simple multiplication math problem under varying levels of difficulty. Some were allowed to use pen and paper, while others had to solve problems mentally. Afterward, participants played a “public goods game” where they decided how much to contribute to a shared fund. Contributions were doubled and redistributed equally, making cooperation beneficial but not individually optimal.
Results showed that when the math task is difficult, and participants received RPI, they contributed more to the group fund, but only when they saw their teammates had similar scores. RPI also increased the time participants spent on the task, reinforcing its motivational effect. In contrast, when performance levels varied or the task was easy, RPI did not increase cooperation.
The second experiment involved 300 working adults who imagined starting a new job at a marketing firm. Participants were told they achieved 12.5% growth in client accounts, while their coworkers had similar results. Some were told they missed a challenging target, while others were told they exceeded an easier one.
Participants were then shown different types of RPI: no feedback, feedback without rankings, and feedback with rankings. Those who received RPI without rankings reported stronger feelings of solidarity, but only when the target was difficult and everyone missed it. Including rankings in the feedback reduced social bonding, even among similarly performing peers.
The study suggests that RPI can be a powerful tool for building team cohesion, but only when it highlights shared challenges rather than competition.
“Providing employees with performance information that reveals common struggles can foster a sense of being ‘in the same boat,” Zhang said. “This shared experience strengthens social bonds and increases willingness to cooperate.”
Zhang believes the findings have practical implications for managers. In environments where collaboration is critical, performance feedback should emphasize shared effort and avoid ranking employees. Further, while difficult goals can naturally lead to bonding, the study cautions against setting challenging targets solely to boost teamwork, as this could lead to unintended consequences.
“When tough goals are already in place, managers can use RPI strategically to reinforce unity and cooperation. My research highlights the risks of emphasizing competition.”
