AI now assembles survey comments, metadata, chats and behavior in one view. Behavioural psychologists say HR must pick outcomes first
AI now handles the employee data analysis that managers once did by hand, which leaves HR to decide what the data should change and who should act on it, from team managers to the CFO, according to workplace psychologists at Qualtrics and Betterworks.
In May, Amazon set a target for more than 80% of its developers to use AI tools every week, then ranked each developer's token consumption – a count of AI use – on internal leaderboards. Amazon workers started gaming the AI leaderboard by running unnecessary AI tasks to raise their counts, before Amazon scrapped the initiative within the same month.
HR teams that track only whether employees use AI tools, rather than whether those tools improve their work, collect the wrong employee data, said Caitlin Collins, Program Strategy Director at Betterworks, a performance management software company in Menlo Park, California. “If we're not accurately capturing data that shows people are driving value within their job ... then we're losing the thread on this," said Collins.
"AI is really good at taking a bunch of disparate data sources and pulling it together very quickly to say, here's what's going on," said Benjamin Granger, a Louisiana-based Chief Workplace Psychologist at experience management software company Qualtrics. "You go do the work to connect with the people, and connect their feedback, back to the action."
What employee data should deliver
Granger said the goal of employee data is behavior change that improves performance, not higher survey scores. "When it's done well, that improves the bottom line, the numbers go higher, the finances get better. The customers are happier and the employees are happier too," he said.
Each level of the business needs different findings from the same data, central to how HR leaders prove strategic value to the business, added Collins. "At the manager level, the data that we're capturing should help them better coach their team and drive higher performance," she said.
For the CEO and CFO, HR should report the approximate return on investment (ROI), the key performance indicators (KPIs) it affects and whether to expand, continue or stop it, Collins said. "I wouldn't leave surface-level activity data up for anybody to interpret and move on with."
Decide what the data is for before collecting more
HR should name the business outcomes the program exists to drive, Collins said, then define the behaviors that show progress toward those outcomes, and identify the day-to-day activities that move those behaviors.
HR should also measure whether senior leaders show the behaviors the program asks of other employees, and use existing burnout and change-readiness surveys, Collins said. Options include the Maslach Burnout Inventory – General Survey and the Readiness for Organizational Change scale, which Daniel Holt and colleagues published in 2007.
Granger said collecting employee data is no longer the hard part. "You've collected the data, because that part's relatively easy," he said. "How do we turn that data into actionable insight? And AI is extraordinarily helpful there."
Who should act on employee data
Collins said HR rarely assigns a specific person to act on what the data shows, or writes an action plan that feeds the results back into the program. HR's role, she said, is to explain what the data means and what each group should do next.
Collins recommends that managers track changes week over week or month over month. "Are we seeing a decline in focus on a project? Are we seeing a decline in collaboration with teams that they might be working with?" she said. Before assuming an employee has disengaged, she asks whether the team has too many assignments and what has changed for that employee.
Granger said AI now matches survey comments to past events and demographics, freeing managers to meet with their teams. That matching, he said, is "relatively less value adding than me actually going to my team and spending an hour talking about, here's what you said, here's the context within which you said it. Let's work on an action plan together and gain some buy-in."
For HR leaders comparing performance management, survey and AI analytics software, that advice raises one question for each vendor: does the system send each finding to a named manager with an action plan, or does it only collect more data? That question matters more as performance management shifts toward frequent feedback.
"Ultimately, I think we need to share information internally to our managers, leaders, executives, because they are accountable for what's happening with that data," Collins said.
