Labor data analytics can inform better talent decisions

Labor data can help organizations make more informed talent decisions, but more companies could be taking advantage of it. 

According to Mercer’s “2020 Global Talent Trends Study,” 39 percent of organizations use predictive analytics to inform their people-related decision making. Thirty-one percent said they use a cause/effect analysis of key workforce and business outcomes, and just 18 percent gather data to assess the impact of different pay strategies on performance.  Meanwhile, 61 percent of executives who have used talent analytics said that doing so to inform decision-making is the number one HR trend that has delivered an impact

A workplace strategy expert spoke to Workforce about the potential of people analytics and how organizations can use it in their talent management strategy.

Using labor data analytics to measure employee impact

People analytics can inform complex predictive models, but it can also help organizations understand  how certain decisions have impacted employees. 

One prime example of this is pay equity, according to Tauseef Rahman, partner in Mercer’s San Francisco office. Labor analytics can be used to understand if there are pay gaps at an individual or group level, after taking into account factors like where employees live, what work they do and how long they’ve been working at the company. 

Also read: Labor analytics add power to workforce management tools

Analytics can also help organizations understand performance measurements better, Rahman said. Based on performance ratings and rewards like promotions or raises, are those performance ratings biased toward a specific subset of employees? 

Doing this analysis at a company-wide level is a good place to start, Rahman said. From there, leaders can see if there’s a company wide gap. There may not be a pay gap within the organization as a whole, for example, but through analytics, one can drill down and see if there are gaps in a specific business unit or a specific team. 

“That data can certainly help shine a light on where those patterns are and then that will help redirect efforts and resources to areas that need the most help,” Rahman said.

Using data analytics predictively 

Remote work has become more commonplace for many organizations during the COVID-19 pandemic, and experts expect remote work to continue to be more widely accepted afterward as well. Rahman said that analytics can help organizations interested in adapting an increasingly remote workforce. 

As organizations consider a more geographically distributed talent pool, many questions arise, like: If we hire people from across the country, how does that change how we do talent acquisition? How does that change how we pay people? And how do we manage the employee experience across different locations?

 Data can help answer questions and  inform decisions like this. 

Also read: The most pressing workforce management issues of 2020

Acknowledge the inherent flaws of data collection

No set of data can be completely unbiased, but what organizations can do to address this fact is simple, Rahman said. The key here is to clearly acknowledge that there’s bias in how data is created. 

Also read: Keeping Data Safe: The Next Wave of HR Tech Innovation

People who use data to inform decisions or strategy can acknowledge this bias by considering a few questions when they plan on using a dataset to do or plan something. These questions include: How was the gathering of this data framed? Was anything missing? Was answering these questions optional or required for survey takers? Were survey creators biased to presume certain outcomes?

For example, in recruiting, Rahman has seen the presumption that no one over the age of 65 would be interested in applying to a tech position. The reasoning behind this may be laziness or mental shortcuts rather than malice or age discrimination, but it’s flawed reasoning no matter the intention. “Things that are done in the spirit of making it easy can result in unintended consequences,” Rahman said.

Cross functional teams are critical

According to Mercer’s “2020 Global Talent Trends Study,” the  “quality and reliability of data are critical.” Cross functional teams are critical for organizations interested in using data the right way, Rahman said. 

The team shouldn’t only include analysts and statisticians but also people with HR expertise, legal expertise and an ethical understanding of data collection. What data does an organization have access to, and how can it collect it in a way that’s not creepy?

“You can configure technology to do whatever you want. If you do something wrong, you can’t blame the technology. You blame the people who configured it,” Rahman said. 

Managing privacy concerns is an important part of these teams. There are interesting ethical questions that come up with the possibility of using labor analytics predictively. For example, Rahman said, what are the ethics of having a model that predicts how likely an employee is to quit? If they haven’t actually quit, what decisions can you ethically make with that prediction? 

Having someone with policy expertise is also beneficial for a cross functional team. For example, if an employee has a 90 percent chance of quitting according to an analytics model, a policy expert could consider what could be driving the employee’s dissatisfaction and what workplace policy could help them be more engaged. Maybe it’s something related to compensation or work-life balance that can be addressed.

The scope of people analytics 

Businesses go through times of uncertainty for many reasons, from global crises to national recessions and more. Times like this highlight the role of labor data analytics to make the employee experience better, at a time when many employees may be going through financial or personal struggles. 

Still, while focusing on improving the employee experience, organizations cannot lose sight of broader, also important areas of business that analytics can inform, Rahman said. These bigger picture topics include pay equity and diversity. 

“Having a broad mindset is really important,” he said. “You want to solve daily issues, weekly issues, monthly issues and multiyear issues, [all while] not losing sight of the fact that you still have to do your pay equity analysis and you still have to make people engaged in your company. It’s a lot more work, but who better to do it than these strong, multi-disciplinary [people analytics] teams?”