Emerging Trend: Four Ways Amazon And Others Are Using AI To Personalize Work Schedules
By Dr. Michael Watson
AI-powered personalization for workforce optimization is the new trend helping companies simultaneously improve worker retention and increase productivity.
It is clear that companies need new methods and tools to help with workforce planning and retention. As of the end of May 2022, the number of job openings in the U.S. was 11.3 million according to the U.S. Bureau of Labor Statistics, and 4.3 million Americans handed in their resignations. I'm not surprised by these numbers. It seems like every manager I’ve talked to lately has mentioned the tight labor market and that even the possibility of a recession won’t make it easier.
An emerging approach to solve this problem is AI-powered personalization for workforce optimization. Personalization has long been used to create new offerings for customers. Now, the same AI techniques are being used to personalize schedules for employees.
I first learned of this trend toward personalization and workforce optimization during a talk by Russell Allgor of Amazon at the INFORMS Business Analytics Conference. Allgor described how Amazon built a system to give workers flexibility in picking their schedules.
I confirmed the trend with Filippo Focacci, Founder and CEO of DecisionBrain. DecisionBrain has been delivering advanced workforce optimization software solutions since 2013. Prior to that, he was doing this for ILOG and IBM.
Here are four ways that companies are using personalization for workforce optimization.
One: Personalization of Schedules at Amazon
Amazon’s fulfillment centers are complicated operations involving in-coming products that must be put away, product returns that must be processed, and the picking, packing, and shipping of customer orders. Lots of other tasks also happen, like shipping between fulfillment centers, gift wrapping, and handling of individual items, cases, and pallets. On top of these activities, demand varies wildly from week to week and throughout the year.
It’s not easy to staff these sites. Like most companies, Amazon needs to retain its most experienced workers for as long as possible and recruit and keep new ones.
Prior to the new system, employees could only pick the day or night shift and what day of the week they started on. If an employee started their work week on a Wednesday, for example, they would work full time until Sunday. From what I could tell, simplicity was the key benefit of this type of scheduling. But it was bad for employees who wanted to work part time or on non-consecutive days. It was even bad for Amazon, since workload changed based on the day of the week.
To create personalized schedules, Amazon needed: a) a way for workers to seamlessly start and finish their shifts at different times throughout the day, b) a portal for employees to enter their preferred shifts, and c) a means of assigning individual employees to different shifts. It’s this last part where AI came in. Amazon implemented an optimization system to assign people as closely as possible to both their desired shifts and total amount of time they wanted to work. The program also balances worker needs with Amazon’s schedule requirements.
In general, this is a great approach that allows companies to better accommodate employees’ desires for overtime, full-time, and part-time work and to recruit people who might require non-standard schedules. By better meeting the needs of employees, companies can improve retention and make it easier to recruit.
Two: Personalized Upskilling
It’s not always clear when demand can be cost-effectively met through upskilling, but some firms are using analytics methods to detect these opportunities and then use the insights to personalize job scope. I know of two firms that employ this method right now – a leading workplace experience and facility management company and an international call center company. Both use workforce optimization software to optimize shift patterns to meet peak demand with existing staff. Sometimes, however, this is not possible. A shortage of skilled technicians at the facility management firm might lead to the recommendation that certain janitors be trained in simple technical jobs, like lightbulb replacement.
For employers, upskilling increases flexibility and smooths workload without requiring additional hires. It can also lead to lower turnover. For employees, it creates a better career path, higher pay and increased job satisfaction, and from a client’s perspective, upskilling improves service. Workforce optimization systems can allow managers to understand and quantify these benefits to help companies make smarter decisions about how much and what type of upskilling they need.
Three: Creating Smart Agendas
A multinational healthcare company uses workforce optimization tools to help their pharmaceutical sales representative (PSRs) create efficient personalized weekly sales call agendas. PSRs must visit physicians at a certain frequency, sometimes virtually but usually in person. These meetings are limited by physicians’ tight visit hours and geographic spread, which results in PSRs spending a lot of time on the road. PSRs must often adjust schedules at the last minute because doctors frequently get called in to tend to patient emergencies. On top of this, PSRs have their own schedule constraints and preferences to consider.
This complexity is what led this company to implement an AI-powered workforce optimization tool that could consider all these factors when creating weekly sales call agendas. The tool has helped their PSRs increase the average number of visits per week and better meet physician visit frequency targets, while relieving sales reps of the burden of manually building schedules.
Four: Better Prioritization
With the vast amount of data now being collected and with better AI forecasting, it is much easier to understand and prioritize work, especially in fast-moving environments that require constant adjustments. Going back to the facility management company I mentioned above, they use workforce optimization to prioritize the cleaning of meeting rooms, only cleaning those that have been used.
Similarly, if a client reports a critical malfunction, the company’s workforce optimization system can automatically adjust the day’s or week’s activities. The new plan might suggest moving some routine maintenance work out to a later day while moving up two maintenance jobs that are located near the site of the high priority repair job. Having visibility into the prioritization of what actually needs to be done means that companies can move away from block scheduling workers and frequency-based tasks and be more reactive to the needs of the moment.
The Trend is just Starting
Workforce optimization is often misunderstood as being about squeezing workers. However, it is not this at all. The examples above show that it can be used to increase employee satisfaction while still helping the company.
In a competitive labor market, the use of AI to personalize the employee experience will only keep growing.
---
About The Author
Dr. Watson is an Adjunct Professor of Optimization at Northwestern teaching optimization and operations management. He is also the lead author of the book Supply Chain Network Design and Managerial Analytics and is an entrepreneur and software executive.