John Sumser: When intelligent tools don’t see what matters

By: | February 12, 2020 • 3 min read
Emerging Intelligence columnist John Sumser is the principal analyst at HRExaminer. He researches the impact of data, analytics, AI and associated ethical issues on the workplace. John works with vendors and HR departments to identify problems, define solutions and clarify the narrative. He can be emailed at hreletters@lrp.com.

Carmen’s device vibrated an alert: Sam’s engagement score dipped below 75 for the third day in a row. The attrition-risk indicator on Sam’s profile was glowing yellow. In another day, it would turn to red. Sam, a key employee and one of her three direct reports, was on his way to becoming an official flight risk.

Carmen’s inbox filled with warnings and advice. “Have you talked with Sam recently?” one email asked pointedly. Another cautioned that “33% of your team is heading for the door.” A third offered advice on ways that good managers improve retention. A fourth delivered an analysis of Sam’s personality, qualifications, engagement scores, promotability index and a comparison of compensation for similar jobs at other companies. The final email offered a list of projects that might be offered as stretch assignments.

Carmen needed to get to the bottom of this in a hurry. If Sam’s engagement score continued downward for five days, she would have to prepare a brief and action plan to report to her supervisor and HR. Sam was being fast-tracked for management jobs and losing him would really hurt the diversity statistics.

Advertisement




Studying Sam’s attendance, Carmen noticed an interesting pattern. Every other week, Sam left early on Thursdays and Fridays. As she was making a note to follow up with Sam on this pattern, her automated coach popped up on her desktop screen. “Lean right into the problem to find your greatest leverage,” it intoned.

Meanwhile, anyone in the department who hadn’t recognized Sam’s accomplishments in the past 10 days was prompted to do so. Various digital coaches and assistants suggested things to do to make Sam feel more included, such as inviting him out for drinks after work.

In Sam’s cube, things were different. Sam was a single dad with two young children, 8 and 10. His mother, who usually handled afternoon childcare, was in the hospital for surgery. He was asking for favors from friends and neighbors to look after the kids in the afternoons.

See also: This is why you have to take employee experience seriously

But, when that didn’t work, his kids were fending for themselves for the first time. It would be four hours between the time the kids got out of school and Sam got home. He normally left early every other Thursday and Friday to get the kids. But he couldn’t pick them up more often because of important project deadlines, and he was low on PTO.

When he found someone to watch the kids, he stopped by the hospital to see his mom on the way home. The hospital wouldn’t let the kids in because of their age. But he never knew when it would work out and he felt guilty for not being there for his mom and his children. Sam felt overwhelmed and trapped.

But Carmen didn’t know any of this. Carmen’s analysis, driven by pointers from her intelligent tools and automated coaches, led her to conclude that Sam’s issues must stem from work. When she talks to Sam, they have a lengthy conversation about a difficult customer and the pressure of some deadlines. She asks how things are going otherwise. Sam says, “Fine.”

Advertisement




Carmen leaves the conversation convinced that the problem will resolve soon. Sam leaves the conversation convinced that the company only cares about its deadlines, not its employees.

Three weeks later, Sam takes a job at a competitor for a slightly lower salary. His new company offers an after-school program, back-up childcare and generous paid-leave policies.

Intelligent tools are not a panacea that will understand all of the real-world problems employees encounter. Our new tools can quantify many things but can only help us see a part of the picture. Retaining Sam depended on his ability to trust Carmen with the details of the real stress in his life. In this story, the program’s analytics and advice limited Carmen’s ability to see a bigger picture.

Related: How VR can inspire better managers

It will be tempting to let our newfound tools take on decision-making. But, the tools can only tell us about the things we measure. They are focused on repeatable events and metrics that can gather and compare data. They cannot anticipate or see the normal, but unpredictable, events that humans regularly encounter.

In my master class at Select HR Tech in June, I’ll be guiding exercises that help clarify the relationship between people analytics, AI and people.