Does this scenario sound familiar? You’re a manager and it’s an ordinary day, nothing exciting is happening but then you receive an email or a meeting request from one of your team members. They inform you that they’re resigning from their position, effective in two weeks. You’re shocked because you didn’t see it coming. There were no indications during your weekly one-on-one. Their work has been good and you thought they were happy. Why did they resign?
“The most common cause,” says Cliff Stevenson, Principal Analyst, HR/Workforce Management at Brandon Hall Group, “is personal conflicts with their supervisors.” These conflicts often aren’t brought to the surface. Instead most people will choose to leave a job.
But what if you could tell when an employee is dissatisfied and ready to move on? That’s where predictive analytics, which takes data, statistical algorithms and machine learning to predict a future event, can help an organization better understand and manage their workforce.
When an employee decides to leave, it’s very hard to stop them or change their mind,
But predictive analytics can help identify high and low performing employees, which can help managers improve personal, team and business performance. It can also help companies attract top talent, and make changes to their hiring, HR policies and company culture to help keep your best workers for longer and look for more top performers.
High performers aren’t always the loudest employees, says Stevenson, who led HR departments before going into research. High performers, based on his experience, are the people who enjoy their work. “They aren’t necessarily completely money motivated. So, when they do make requests, they’ve very polite. They’re not aggressive.” That attitude can be detrimental he says, because companies may be expecting them to speak up when they may not be comfortable doing so.
It’s situations like these that explain why companies often don’t recognize and reward high performers until it’s too late and that top employee has mentally and physically checked out of the job. That can cost an organization a significant amount of money.
Replacing employees affects an organization’s bottom line. A report from the Center for American Progress found that replacing an employee can cost 16% to 21% of the position’s salary, and the more specialized the position, the higher the percent of the salary. So, if your organization has high employee turnover, your bottom line could be taking quite a significant hit.
Replacing an employee also takes time including:
This process could take a couple years for a new employee. Therefore, companies are looking at predictive analytics to reduce the turmoil of turnover. It’s cheaper to keep an employee than it is to find and train a new one.
With the right data, predictive analytics provides a strong advantage.
Each organization is different, but analytics can help target applicants, prioritizing the most qualified and the one who would best fit in your organization’s culture; help determine how to best fill a position; identify factors that create employee satisfaction so they can be replicated and improved upon; identify reasons for workforce attrition instead of leaving it to an educated guess, and finally, help organizations create effective training initiatives that are targeted and useful.
It all starts with data, says Stevenson. “Rather than looking at one particular data point, [predictive analytics] looks at how different data points relate to each other. So, somebody being late or missing a few days of work is of itself not going to tell you anything. It might be something happening in their life.” What the data does do is take those different data points and identify trends.
“When they see a certain set of parameters that fit a pattern, they know that in a certain percent of cases that person has left within three months,” says Stevenson. This doesn’t mean that every time a manager sees this pattern that their employee will leave, but there is a chance that they might. Predictive analytics lets a manager know that they might want to take a stronger look at that employee.
Can predictive analytics stop a high-value employee from walking out the door? It can, says Stevenson, but it depends on two things.
“It depends on how much data we have on that employee and for that organization. Sometimes it takes a while to have enough data to find a pattern. In that case, you do run the risk of it being too late.”
The organization also plays a role. “The other part of it is: what is the underlying change that you can make that would cause them to stay,” he says. “If it’s just a matter of compensation or you know the person is looking for some internal mobility, as long as there is enough time to reach that person with the right understanding of their motivation, you don’t need much to get them to stay.”