Erin Kelly
PHOTO: ISTOCK / STOCKROCKET
In addition to making modern life run more smoothly, Kelly observes, fulfillment centers offer something else of high value in the American economy: jobs with decent wages that require neither college degrees nor specialized educational credentials. “For noncollege workers,” she says, “they’re one of the more consistent, regular, and higher-earning options.”
Unlike restaurants and retail stores, where employees often struggle to get adequate hours and predictable schedules, fulfillment centers offer full-time hours with benefits and overtime possibilities. “So this sector is very interesting in that it is providing a possibility for noncollege workers to have a steady and slightly higher income than they could in the service sector,” Kelly points out.
To be sure, this is challenging, tightly monitored, and stressful work. “It’s really good work if you can get it,” Kelly says, “yet also really tough work. We know that this is a growing sector and our research team wanted to see whether job quality could be improved in a way that would benefit the workforce.”
Pride and pressure
Beginning a research partnership with a major corporation (not named to protect confidentiality) in 2019, the team interviewed more than 100 frontline workers and managers in multiple fulfillment center sites.
“We heard about physical exhaustion and feeling that they’re under time pressure, but also a sense of pride in the ability to handle the pace needed to meet this goal of getting the products on the trucks every night,” Kelly recalls. “Some expressed appreciation of working in a culture where productivity was valued and liked that it was not a place for loafing.”
While the employer had concerns about employee turnover, Kelly says, their partnership with MIT Sloan was driven more by a desire “to explore how to respond to the constraints of needing to do next-day delivery and the operational realities of e-commerce, yet look for a way that would be consistent with the company’s stated values and supportive of frontline workers.”
Post-2020, a new reality
When the pandemic hit in February 2020, Kelly’s team paused their research for nearly a year. Once they resumed, consumer demands had skyrocketed and the pressure on fulfillment center workers had intensified. Kelly and the team had previously planned to address scheduling policies, but in light of the broader concerns raised in this pandemic period, “We ended up designing—in dialogue with them—what we called a new health and well-being committee (HaWC) that we tested and evaluated with a field experiment.”
The HaWCs are small groups of employees across different departments, including one or two supervisors, that meet regularly and welcome employee concerns, suggestions, and identification of problems. “It’s framed as being a chance for people to share what causes them stress at work,” says Kelly. “Often those end up being operational concerns that also cause frustrations for workers; people would share things like, ‘There’s no broom in this part of the building, and so I have to take time to go get one, and then I have to work faster for the next portion of my shift.’”
Kelly notes that it is critical for HaWCs to be centered on workers’ needs rather than on employer goals. The researchers are also keenly aware of mental health challenges in this population of workers. “Over half of our survey respondents were at moderate or severe psychological distress. Moderate distress is not necessarily clinical depression or another clinical diagnosis, but it is something concerning enough that practitioners would recommend taking some action.”
Management response to worker requests raised by the HaWCs, of course, is vital to their mission and legitimacy, and respond they did. “They made a wide variety of changes,” says Kelly. There were simple fixes like adding brooms where employees had highlighted a need, but also new ideas like playing music in the workplace and giving everyone input into the playlist. “They set up systems where people could either vote for specific radio stations or contribute top-10 songs that would go on the building’s new playlist. That was a smart strategy because it was relatively easy to implement and people would notice it.”
More structural changes made because of HaWC feedback related to training, both of new employees and for those asked to cover for absent workers in other departments. In one fulfillment center, says Kelly, the HaWC recognized that new employees were struggling. “They proposed to management that the training period be extended by another couple of weeks and that people be given a department buddy, another coworker to support them.” For workers stressed out when they were moved to an unfamiliar department, a group created cheat sheets—visual guides to tasks. “The committee was able to think through, ‘What is the problem here and how could we address it?’” Kelly recalls.
Encouraging results
The team was pleased to find that after six months, employees in the buildings with HaWCs were 33% less likely to report moderate or severe psychological distress than employees in the other buildings. “We find short-term benefits for mental health,” says Kelly, “but this is still exciting because we see it is feasible to support these workers’ well-being by making changes in the workplace.” The researchers also found a striking reduction in employee turnover, which was welcome news for the employer, since high turnover results in higher training costs and lower productivity.
“We rolled out the HaWCs in eight buildings, and we had eight buildings that served as our control group and continued on with their usual practices. We have strong experimental evidence that this reduced turnover significantly,” Kelly reports. “Employees in the buildings with a HaWC had a 20% lower probability of exit as compared to the previous year’s exit rates. What we’re seeing is that having a chance to share your concerns and have a voice in this very regimented, technologically monitored setting—that’s affecting people.”
While workers serving on the committees generally valued that experience, the analysis focuses on impact across the fulfillment center, including many workers who were not directly involved in the HaWCs. “There are spillover effects that we think are indicative of people feeling respected. Seeing your peers be actively involved in problem-solving—that signals that the organization appreciates the perspective and wisdom of frontline workers,” Kelly says.
Human solutions more powerful than technology
Another aspect of the research was the opportunity to compare the employee-driven HaWCs with a technological approach. The company introduced kiosks in break rooms of many of the same fulfillment centers and invited employees to use them to identify workplace problems. Workers could type their feedback and concerns privately and were assured that responses would be communicated directly to management. But the team found that while the HaWCs, which offered a worker-led, community-building voice channel, significantly reduced turnover, the kiosk had no impact at all.
Kelly’s team has presented their findings to the company, and will monitor longer-term results. They are also submitting several academic papers, and Kelly has written a teaching case that she presented to participants in the MIT Sloan Fellows MBA program. Her hope is that the team’s findings will lead to improved well-being for workers in this growing industry.
The influence of the HaWCs on employee decisions of whether to stay at or leave their jobs, Kelly says, “was intriguing evidence. The human element here is the signal of being respected, of being heard, of being valued as a member of the organization, beyond whether you complete your specific tasks.”