Welcome
Payroll News

Robots, Artificial Intelligence Are in Payroll’s Future

June 3, 2020, 9:02 PM

Robots and artificial intelligence are not just sci-fi movie stars. Payroll is moving into the cyberworld as a way to improve efficiency, two executives said June 3.

The use of robotic process automation (RPA), artificial intelligence, and data analytics can help employers streamline payroll processing, analyze information, and reduce costs by having computer software perform tasks that often are conducted manually, said Joe Ziska, manager of compensation and organizational management at the automaker BMW Manufacturing Co.

Employers’ needs are rapidly evolving, especially during the novel coronavirus pandemic, Ziska said. The number of teleworkers in the U.S. rose to 64% from 29% since mid-March, and gig workers make up 36% of the workforce, either as contractors or as members of a temporary workforce, he said.

“Workers suddenly needed to receive payments while working remotely, in many cases for the first time,” Ziska said. Companies are struggling to keep up, and with 30% of businesses using payroll systems that are at least 10 years old, the move toward automation to improve efficiency has employers open to new technology, he said at the annual American Payroll Association Congress, which was held online because of social-distancing requirements from the coronavirus crisis.

Automated Robot: Not a Machine

The use of RPAs is one solution to improve efficiency, said Fidelma McGuirk founder and CEO of Payslip, a company in Ireland that provides technology to help companies automate payroll processes and standardize data. Robotics can be used to “take over standard and repetitive activities that are currently carried out by humans,” she said. “It increases the value of the work done by the human by removing work than can be repetitive.”

“An RPA is not a physical machine,” McGuirk said, unlike the sci-fi machines from movies and novels. In payroll, an RPA would not replace existing payroll software, but could be used in support of data-driven tasks, such as payroll validation and reconciliation. If errors or inconsistencies are found, the RPA can return the file for a manual review, she said.

Additionally, with so many employees teleworking, distractions at home would be negated with the use of RPAs because automation would reduce the chance for errors, McGuirk said.

While RPAs can take over manual tasks, the software cannot learn.

Artificial Intelligence

“RPA is a technology that follows predetermined rules to help automate tasks. It’s elegant but it’s also static, and it can’t learn,” Ziska said. “AI commonly describes technology that has the capacity to get smarter over time. The term for that is machine learning.” As machines get smarter, their capabilities increase, he said.

Common uses of AI in payroll include self-service support, reporting requests, and technical troubleshooting, Ziska said. An added benefit: Bots can run for 24 hours.

In choosing to adopt AI in the payroll process, Ziska said employers need to be flexible and open to change, prepare for a long-term commitment, react quickly to unexpected developments, and perform due diligence with the AI service provider.

“AI always gets billed as not needing to take lunch breaks or being able to work overtime for free,” Ziska said. “In the same light, you also invest as much time up front as you would a person.”

Analytics as an Alternative


If an organization is not ready for RPA and AI, data analytics is an alternative.

Analytics can help employers sort out data, such as how much overtime was incurred in a month during the pandemic or the difference between increase payroll costs vs. the number of employees. In the process, large amounts of information from various sources can be analyzed to uncover hidden patterns and unknown correlations, McGuirk said.

An analysis of payroll data might show unexpected upward or downward trends, resource costs by location and specific job roles, and measurements regarding pay and total costs, McGuirk said.

Through analytics, employers can focus on a large population of data rather than a sample, which greater awareness of the information behind the figures, McGuirk said.

You get more insight into what’s happening within the figures,” McGuirk said. “Then you can make more informed decisions.”

To contact the reporter on this story: Michael Trimarchi in Washington at mtrimarchi@bloombergindustry.com

To contact the editor on this story: Howard Perlman in Washington at hperlman@bloombergindustry.com

To read more articles log in. To learn more about a subscription click here.