Harnessing the 'Spillover' Effect
New research finds bottom-line benefits when organizations spend time carefully considering which employees should sit near each other. But experts warn that the "peer spillover" effect can also have negative consequences.
By Maura C. Ciccarelli
What if improving employee performance by nearly 15 percent was as easy as adjusting seating charts?
That is the intriguing conclusion reached recently by a team of researchers from Harvard Business School in Boston, hiQ Labs, a San Francisco-based talent analytics firm, and Cornerstone OnDemand, a global cloud-based learning and talent management software company headquartered in Santa Monica, Calif.
In recent years, organizational psychology research has found that employee productivity and quality improve because of a "peer spillover" effect; that is, top performers have a positive influence on people who aren't as highly accomplished.
To see how proximity affects such spillovers, the research team -- Dylan Minor, assistant professor at Harvard Business School, Michael Housman, a workforce scientist in residence at hiQ Labs, and Yitzi Greenbaum, Cornerstone OnDemand data scientist -- looked at two years of data from a large technology company that had performance measures for some 2,000 employees located in four call centers (two in the United States and two in Europe), as well as seating charts.
First, they classified call-center employees in three categories: highly productive workers (who were low in quality), high-quality workers (who were low in productivity) and generalists, who had average productivity and quality scores.
Then, the seating charts and measuring tape came out. They correlated the employees' locations vis-a-vis other workers and saw where performance improvement "spillovers" occurred. They found that pairing highly productive workers with high-quality workers and seating generalists with other generalists generated up to 15 percent of increased organizational performance, which represents the average of a 13-percent improvement in productivity coupled with a 17-percent gain in effectiveness.
Translated into dollars and cents, the study predicts a boost of $1million in productivity per 2,000 employees as a result of these pairings.
The researchers found that the improved performance occurred almost immediately when workers with complementary strengths were seated next to each other and vanished within two months of exposure. It suggests that the source of the spillover effect is a combination of inspiration and peer pressure from being in close physical proximity.
Spillover effects create opportunities to pair people so that they have a better aggregate performance than if the seating chart was left alone, Housman says. "You got better as long as you sat next to someone who was a strong performer, but as soon as you moved away, your performance went back to its original level," he says. "We can't rule out learning, but we think this effect was [related to] either competition or inspiration. If you sit next to someone who is a really hard worker, you say, 'I'd better step up my game,' and you tend to do better, but as soon as you move way, you say, 'I can take it easy and take a breather.' "
Jason Corsello, senior vice president of strategy and corporate development for Cornerstone OnDemand, says his company began investing in the data analytics side of its enterprise software business about three years ago. That's when it acquired Evolv, a company with a data science and machine-learning platform that could be used to draw conclusions and make predictions from the vast amount of data collected on the company's 30 million users. ("Machine learning" is a type of artificial intelligence that lets computer programs change and grow when exposed to new data.)
Corsello said the study's results were shared with the client to help them understand how seating affected performance.
"Many companies have a lot of interesting information," Corsello adds. "Part of the work we do with clients on big data efforts is to manage the aggregation and anonymization of the data."
How can companies apply this learning to their own organizations? Corsello suggests that companies track employee performance over time relative to their seating assignments and proximity to high-performing co-workers. "You can't change an organization overnight," he says, but you can "see a change in a company over a year's time."
There were also conclusions related to negative spillovers. By reviewing records of employees who were fired because of misconduct or unethical behavior, the researchers suggested that, when adjusted for other factors such as work environment and trust in co-workers and managers, workers may be 27 percent more likely to engage in toxic behavior when exposed to another toxic worker.
Warning signs for toxicity can be found in employee engagement surveys, enabling a company to take earlier action. The study showed that, when a toxic worker is removed from a team, the adverse effect on the rest of the team has largely dissipated within a month.
The research is the latest showing the profitability of strategic application of organizational psychology, says Seymour Adler, a partner at Aon Hewitt and an adjunct professor at Hofstra University's doctoral program in applied organizational psychology.
"What we are seeing here, and what organizations are taking advantage of, are the basic social influence processes," Adler says. When people join an organization, he adds, they have to learn by observing their co-workers and inferring what the culture requires for top performance. That can be tricky when toxic behavior is front and center in their immediate environment.
"That's why putting people near people who are high performers can help improve the newcomer's game," says Adler.
He also points to the concepts of social evaluation -- i.e., being judged by your peers -- and positive social contagion -- i.e., doing things that your co-workers are doing -- leads to better performance when top performers sit nearby.
Companies such as General Electric, for example, are changing their performance appraisal systems to capitalize on the power of social evaluation to encourage and improve performance. GE just announced that it was making its new app, which lets employees rate each other's performance anytime and anywhere, the replacement for once-a-year formals review by managers. "Millennials, in particular, are sensitive to [such] social evaluation," Adler says.
What are the next trends in office design and productivity?
Stacia Sherman Garr, vice president and talent management research leader at Bersin by Deloitte, says the recent open office trend is a way to spur interaction between people of different jobs and departments to cross-fertilize ideas. But it's not a panacea.
"It is important to have incidental interactions in a world that is increasingly virtual, but I don't know that it has to be due to physical proximity," Garr says. "I think it is [more about] network or relationship closeness. Some of the research [shows] that those individuals who have more close networks tend to bring a greater diversity of ideas, and that pushes innovation forward."
Send questions or comments about this story to firstname.lastname@example.org.