zagandesign - Fotolia

Employee turnover prediction takes aim at rising quit rate

The economy is encouraging employees to look for new jobs, and the quit rates are rising. HR has a variety of tools and methods for detecting turnover risk.

Thanks to the economy, firms have good reason to worry about employee turnover. But HR managers have access to a range of tools that claim to identify turnover risks. They can spot behavioral clues in employee communications, glean sentiment in ongoing surveys and analyze pay, benefits and even commute time.

The employee retention problem is growing as the economy improves. The number of people who voluntarily left or quit their jobs in June alone was at 3.4 million. This year has levels of quits per month not seen since about 2001, according to government data. And it gets worse for HR managers. A recent survey by OfficeTeam, a Robert Half firm, found that two out of five workers would leave their jobs for better pay.

Vendors are offering what amounts to an early warning system that HR can use to flag employee turnover risks. But vendors aren't promising turnover reductions from the use of these tools. How a firm responds to the issues that prompt employees to leave, whether a bad manager or low pay, is a separate issue. The tools are intended to help identify the risk factors.

The big data approach to turnover risk analysis

ADP, for instance, uses big data -- anonymized data from thousands of client firms -- to run a predictive analysis of employee turnover risk.

ADP can run an industry-wide comparison of pay and bonuses of people doing similar jobs. It can look at how long someone has been in a job since a prior promotion and how that compares with peers. Another clue to turnover is the length of an employee's commute. The longer the commute, the greater the risk.

[Employees] have more choice and are picking up and taking their talents elsewhere.
Marc Rindchief data scientist, ADP

The size of the team may also be an indicator: If an employee works as part of a larger team -- 20 or more, for instance --  a company will see more churn, according to Marc Rind, chief data scientist at ADP. The number of absentee days and whether employees are using benefits also can be risk flags among many other signs, he said.

Employees today "have more choice and are picking up and taking their talents elsewhere," Rind said. There is also a lot more transparency today in the labor market, he added, and employees may know about the benefits and time to promotion at other employers.

Monitoring employee communications for clues

Another approach to detecting employee turnover is used by Wiretap, which can monitor workplace communications on platforms such as Workplace by Facebook, Yammer, Microsoft Teams, as well as its Groups platform.

The Wiretap monitoring can track employee sentiment and assess negative or positive conversations.

If there is negative sentiment associated with a particular manager, the firm can act on it and perhaps coach the manager, said Jason Morgan, Wiretap's vice president of behavioral intelligence.

The system can be used to flag, for instance, harassing behavior, which could lead to employee turnover among other workplace problems. It can drill into private messages to try to uncover the behavior of individuals, Morgan said.

The role of engagement analysis

Engagement surveys may also act as a warning system on employee turnover.

For example, Qualtrics' platform measures four things: employee, customer, product and brand experience. "All four are highly correlated to one another," said Mike Maughan, head of global insights at Qualtrics. "If your product is terrible, your customers are always mad," and this means your employees have a "bad experience." Hence, the potential for higher turnover.

In terms of employee experience measures, a firm can already tell what its turnover rate is from operational data, Maughan said. Qualtrics works to discover the cause of turnover partly through employee engagement surveys, including pulse surveys, but also though an analysis of operational data coupled with macroeconomics conditions that may make it harder to retain employees. Qualtrics develops predictive models from this data.

Qualtrics' employee surveys seek to find out a range of things, such as manager effectiveness, an employees' sense of alignment with the business and whether they feel their job is contributing to the firm.

A goal is to understand the root causes behind turnover. Employers can't assume that they know, Maughan said.

"Maybe the reason people are leaving is not because of higher pay, but because they want, say, flexible work hours," Maughan said.

More tools are arriving. The forthcoming LinkedIn Talent Insights platform will provide turnover rates at various firms.

LinkedIn's tool will monitor changes in user profiles and use that to calculate attrition rates. It claims it will be break down attrition rates by skill sets, allowing a firm to compare its attrition rate in one occupation with those of its peer firms. Talent Insights is due to be released this fall.

Dig Deeper on Core HR administration technology

Business Analytics
Content Management
and ESG