Jacob here. In 2011, Hewlett-Packard tested its people-analytics chops by developing a model to predict employee turnover. According to computer scientist Eric Siegel’s book Predictive Analytics: The Power To Predict Who Will Click, Buy, Lie Or Die, the model allowed HP to assign every individual employee at its company a “flight risk” score that represented the probability that that employee would leave the organization.

Companies can use a variety of methods to predict employee turnover. HP, for example, used two years of employee-level data on things like salaries, raises, job ratings, and job rotations, according to Siegel’s Predictive Analytics. Other organizations may use data from employee-engagement surveys for workers who have stayed and left. Both of these approaches rely on collecting and combining data on individual employees.

This type of analysis is “pretty easy to do,” says Michael Lieberman, a statistician and the founder of statistical consulting and market research firm Multivariate Solutions. And with a tight labor market, it can be tempting for companies to develop this type of model, predicting which individual employees are at risk of leaving. The question is whether the benefits outweigh the privacy and ethical concerns, which could ultimately harm your organization.

“The scientist in me is like, ‘That's so cool. How useful is that clarity?’” says Dr. Kenneth Matos, director of people science at employee-feedback platform Culture Amp. “The ethicist in me is like, ‘Wow, there's so much opportunity for that to be abused.’”

We often write about different tools that can make life easier for HR professionals. Today, we’re focusing on one that seems to do just that, but comes with hidden pitfalls. Here are some of the drawbacks of using predictive analytics to forecast turnover at the individual level:


What is your biggest challenge?

We can help, with Charter Pro research, reporting, and advisory support.

What's included:

  • Access to on-demand advisory from Charter team & experts
  • Data-driven research and insights
  • Actionable journalism, toolkits, playbooks
  • Private roundtable sessions