The hidden cost of just-in-time scheduling

Do you know your work schedule for tomorrow? How about today? For millions of Americans in the service industry, the answer is no.

Employers are increasingly using “just-in-time” scheduling to more closely match front-line staffing with unpredictable customer demand. Such flexible scheduling provides a far more efficient use of labor than traditional fixed scheduling allows.

But there are some hidden costs, as identified in a new study by Yong-Pin Zhou, a professor of operations management at the University of Washington Foster School of Business, and co-authors Masoud Kamalahmadi and Qiuping Yu of Indiana University.

The research trio finds that the most extreme form of just-in-time scheduling—requiring service workers to extend shifts during busier than expected periods—can reduce sales.

“We show that short-notice schedules, issued a day or two in advance, do not harm server productivity compared to regular schedules,” Zhou says. “But real-time schedules, which offer no advance notice, hurt server productivity by 4.4 percent.”

Flexible or unpredictable?

In the service industry, labor accounts for as much as 70 percent of a firm’s operating costs, yet it’s notoriously difficult to get the right number of employees working at the right time due to demand uncertainty and advance or fixed scheduling.

Under-staffing leads to bad customer experience and the loss of sales. But over-staffing leads to unnecessary labor costs.

Enter just-in-time scheduling. Today, service firms often set a base staffing level in advance to cover the most likely demand scenarios. Then they adjust the number of employees and their work hours on the fly based on customer turnout.

An estimated 17 percent of the American labor force works schedules that are unpredictable or unstable. This includes restaurant servers, retail cashiers, delivery drivers, call center reps and many other types of customer-facing workers.

And while employers love the flexibility of just-in-time-schedules, their front-line employees generally hate the uncertainty they wreak, rendering it difficult to schedule child care, pet care, appointments, social engagements and many other personal and family matters.

A drag on sales

But does this uncertainty take a toll on the bottom line?

To find out, Zhou, Kamalahmadi and Yu analyzed 1.5 million transactions at 25 full-service casual-dining restaurants in a national chain. They were able to observe the pre-scheduled shifts, actual hours worked and sales performance of each server.

The instances of flexible scheduling were bisected into two distinct types:

• Short-notice schedules, set at least a day ahead. So, for example, knowing that a nearby college basketball game will end at 10 p.m. tomorrow, a restaurant might call up additional servers to meet the anticipated rush.
• Real-time scheduling, which require servers to extend their shift beyond schedule to meet an unexpected influx of customers from, say, a serendipitous tour bus arrival.

To isolate the effect of scheduling on server performance, the researchers carefully controlled for everything from fatigue to general busyness to number of tables served.

And using this apples-to-apples analysis, they found that short-notice schedules had no discernible impact on the servers’ sales. Real-time schedules, however, reduced sales by 4.4 percent.

“In light of the restaurant industry’s net profit margin as low as two percent,” Zhou says, “a 4.4 percent loss in sales is surely economically significant.”

A drain on effort

He attributes the dip in sales under real-time scheduling to a drain on efforts to up-sell and cross-sell. An example of up-selling is when a server pushes more expensive menu items, such as recommending a nice glass of wine that pairs better with a meal than the water or soda initially ordered. An example of cross-selling is when a server pushes complimentary items, such as highlighting those amazing fries that you simply have to try with that burger or offering a taste of dessert to finish off a satisfying supper.

This reduction in sales, Zhou adds, is not necessarily due to any conscious sandbagging: “In our study, the servers didn’t realize they were failing to make the extra effort. This can only be observed statistically through the data.”

The study also indicated that this reduction in sales under real-time scheduling was more severe for servers with less inherent sales ability and during weekends.

Mitigating uncertainty costs

Just-in-time staffing schemes provide clear and significant cost-benefits to firms in every service industry.

But this new study points out some downsides, too.

Beyond the hidden cost of reduced sales, Zhou also suggests that the uncertainty resulting from just-in-time scheduling likely leads to increased employee turnover. “If you can give employees a little more stability and predictability—even it’s on short notice—you may avoid a dip in sales and even reduce voluntary turnover in the long run,” he says. “We don’t have data on this yet, but it could result in major savings on hiring and training.”

Beyond the hidden tolls of just-in-time scheduling, an increasing roster of cities—including Seattle—is moving toward an explicit toll on employers by enacting laws that require higher hourly pay for employees working on just-in-time schedules.

Zhou and his co-authors have developed a model that accounts for potential costs added by these new predictable scheduling laws. “In this case,” he says, “we find that a shift away from the heavy use of real-time schedules toward short-notice schedules can improve a firm’s profitability by up to 4.5 percent.”

That’s serious currency in the notoriously tight margins of service.

“Call to Duty: Just-in-Time Scheduling in a Restaurant Chain” is the work of Masoud Kamalahmadi, Qiuping Yu and Yong-Pin Zhou.