Using AI and Data to Improve Employee Engagement

​In 2015, Microsoft began gathering data about its employees to see if it could use the information to make changes in human resource management. In early 2018, the data showed widespread dissatisfaction among the 700 or so employees in one division. These workers expressed more dissatisfaction than comparable divisions at the behemoth software company, according to The New York Times.

Much was at stake—the unhappy employees included engineers with specialized skills and who would be difficult to replace. Microsoft had to figure out why employees were unhappy. So it sought data to explain the reasons for the dissatisfaction.

Harnessing Happiness with Artificial Intelligence

Microsoft, with the help of a data analytics firm, processed volumes of data. The company examined employee calendars and the frequency of meetings. It could tell that employees spent an average of 27 hours a week in meetings, often with a whopping 10 to 20 people at each gathering. Moreover, although the data showed that workers frequently performed well after transferring to different divisions, Microsoft had barred employees from such transfers unless they'd been at their current division for 18 months and a supervisor approved the transfer.
E-mail use was analyzed and revealed just how shackled workers were to this communication medium—how long they spent reviewing and replying to e-mails and at what times of the day. The results signaled that workers spent long hours on these tasks.

After the company reviewed the data, its approach to meetings changed, new policies encouraged time away from e-mail, and employees could transfer to different divisions more quickly.
Data analysis seemed to work. 

"This data collection allows supervisors to better understand areas ... for process improvements, including in areas of added machinery, new software updates or even the addition of new employees," said Mitch Collier, vice president of product management for StayWell in Portland, Ore. "From our experiences, the use of AI [artificial intelligence] tools on work performance has been a positive effort to protect the health of employees and to find ways to help employees be more productive and effective in their roles."

Brad Goldoor, the chief people officer at Phenom People in Ambler, Pa., uses digital tools to find talent for organizations. Through analyzing employees' interactions and activity on an AI-driven internal platform, he said, employers can better assess employee engagement. AI tools can tell managers, for example, which employees in a department seek to move to another or how many employees are looking for a career change.

Big Brother Syndrome

When employers utilize AI monitoring, however, they should avoid making employees feel as if they're constantly being monitored.

Such scrutiny can have negative effects if leaders don't clearly explain its purpose, said Ingrid Fulmer, associate dean and professor in the Rutgers University School of Management and Labor Relations in Piscataway, N.J.
"I think one important key to effectively using AI to monitor performance has to do with understanding employee perceptions of the organization's intent," she said. "The more AI is perceived as being used with the intention of catching employees doing bad things—keystroke monitoring to curtailing Internet surfing—rather than for more positive reasons, such as improving productivity or making work more interesting, the more likely it is to be perceived as intrusive and viewed negatively or with resentment."

[SHRM members-only toolkit: Managing Workplace Monitoring and Surveillance]

Discovering Best Practices

When AI tools lead to improvements for the workforce, employees accept these tools as a means to find better practices—for them and the entire organization.

"Given the large amounts of data … being created daily, based on all the interactions that the employees are involved in, as well as all of the transactions that employees initiate around their schedules, the use of artificial intelligence is the only way to make heads or tails of the data in a meaningful way," said Rich Correia, director of product marketing at NICE Systems in Richardson, Texas.

"AI is constantly looking and working through the data to spot unique characteristics that have a predictable pattern or trend," he said. "The AI discoveries provide insight on what behaviors are helping or negatively impacting results. That same information is also being shown to the managers, along with a second level of AI insight on the best practices for coaching the employees to achieve desired goals."

Jim Romeo is a technology writer focused on business and technology topics.