Internet mayhem broke out last month when the Wall Street Journal published a report on how employers are using outside firms to predict health risks in their employees. Journalists and employees alike expressed their concerns about the sensitive health information employers might have access to and how they’re using it.
Would employers use data about their workers' health habits to predict who would become sick or pregnant or develop serious health conditions? And would they then use this information to terminate these employees, to cut down on healthcare costs? This debate will continue to be heated. But HR analytics -- used respectfully, with a partner you trust -- are important and can be helpful in guiding talent management and hiring decisions. What's more, analytics isn't going anywhere any time soon. According to a 2015 report by Deloitte, 35 percent of companies surveyed said they were actively developing data-analysis capabilities for HR.
From the big-picture perspective, of course, employers are in an awkward position: Should they use HR analytics? And if they do, how can they use this data in a way that doesn’t upset employees or violate privacy and HR laws? Here are a few types of data to carefully manage when you're conducting HR analytics, as well as some best practices for staying compliant with the law and honest with employees:
1. Health data
With the focus on healthcare costs and wellness programs, many employers are now using HR analytics to evaluate their programs' effectiveness, determine gaps in healthcare coverage or employee benefits and help improve programs overall. Using employee and company data in conjunction with industry data can help you find the best benefit offerings and get employees great, affordable care; but employers need to be very conscious of and sensitive to employee privacy at the same time.
The biggest issue is a potential violation of HIPPA or other employee privacy law. Here, employees need to opt-in, to have their data collected and used. Then, there must be enough data generated by enough employees that employers can’t pinpoint which data belongs to which employee. When the data is analyzed, that process should be done in a way that looks at the organization as a whole, rather than individual employees and subgroups.
2. Predictive-performance analytics
For example, the people-analytics team at Google actually developed a formula to make promotion decisions and found it 90 percent accurate in pointing the company toward the right decisions. But the formula was never used. Why? Analytics can give HR the information needed to make better decisions but shouldn’t replace the decisions people make with algorithms. When employers use predictive models to decide not to train people who, for instance, are on the verge of being either fired or awarded promotions, they’re basing their decisions on what an algorithm says may or may not happen, rather than what employees are actually doing. People are unpredictable, and unknown factors can influence outcomes. Decisions that affect people should be informed by data, but made by people.
Although predictive performance analytics seems like a proactive step to weed out employees who will eventually be assessed as delivering a poor performance, it’s a recipe for a wrongful termination suit -- not to mention an action that's unfair to employees. Here’s another example from Google: Back in 2011, its people-analytics team used data to determine the qualities of the best and worst managers. But the team didn’t use the resulting insights to actually get rid of managers displaying those "worst" traits. Instead, the insights were used to improve performance. The message: Be clear about which data from internal sources will be used for what purposes; and use it to help improve management and performance, not to weed out the weakest link.
Analyzing irrelevant information
When employers look at HR analytics, they may get lost in the small details and focus on the wrong things. For example, analytics may show HR that the closer, geographically, that employees live to the office, the less likely they are to leave than those who live farther away. But does that mean the company should pass up on top talent just because those peoplle live more than a half hour from the office? Probably not.
Using specific data in this way is legal, but is it ethical? Can people be fairly evaluated on actions they haven’t yet done, but data predicts they may do? Will using data in this way lead to the best decisions for the business?
When analyzing data, look at the big picture rather than zeroing-in on small minutiae -- such as the fact that one employee took more days off than another, for example. Always focus on making the best decisions for employees and the business as whole.