A new employee is one of the most expensive things there is in business. The hiring process means lots of man hours in reviewing applications and conducting job interviews. Once the new person is hired, they have to be trained to do their job. They will make mistakes that cost the company time, which is money, and money, which is also money. Compounding the hiring problem is the first year employee attrition rate, which has been on an upward trend over the last few years. According to PwC Saratoga’s “Human Capital Report,” first-year turnover has increased from its historic low of 21.5 percent in 2011 to 22.6 percent in 2012 to 24.1 percent in 2013. While the attrition rate typically goes down after the first year, the cost to replace an employee stays significant. According to a report from NOBSCOT Corporation, a company that develops human resources software, “the typical rate of voluntary turnover for many companies is around 16.6 percent per year, which means that most employees stay at their organizations for an average of six years. At the 16.6 percent rate, companies with 2,500 employees lose 415 employees per year and have to spend more than $2 million dollars each year (using the average cost of replacement at a conservative $5,000 per termination) to replace those employees.”
What makes things worse is that many younger workers that are joining the workforce today don’t feel the need to stay with a company if it isn’t the right fit. The hiring process remains the lynchpin that could save companies, and the American economy, millions of dollars. Fortunately, like so many other things in our lives that have been made better by technology, it can also be our savior in the hiring process.
Let’s play a game
Predictive hiring is using data and algorithms to match job candidates with the right job. Mercer, a global consulting leader in advancing health, wealth, and careers, recently made an investment in Pymetrics, a 3-year-old firm dedicated to advancing the hiring process by using their unique approach to predictive hiring. “Pymetrics’ approach to predictive hiring is using neuroscience, big data and algorithms to match candidates with the right job. A lot of companies use some type of selection tests to screen candidates, what Pymetrics does is looks at 80 cognitive traits of top performing employees in a role,” said Barbara Marder, a Senior Partner with Mercer.
To get these 80 cognitive traits seems like a monumental task and might even invoke images of hooking up the perfect employee to a towering supercomputer to scan their brainwaves, or perhaps a mad scientist cackling non-stop as he takes blood samples from an unwilling, yet brilliant, top employee. In reality, it’s not that dramatic. In fact, as Marder described, it’s sounds pretty fun. “A company would have their top performers play a neuroscience game and the data behind it creates a predictive model based on their traits. Job candidates play the same game and the trait profile of the top performers is then used to look through any number of candidates to compare their game performance to the top performing game performance,” Marder said. “These games are grounded in decades of academic research in neuroscience, to tell you about people’s natural traits and attributes.”
Many times, if a job applicant knows that there will be a game involved in their hiring, then they might try to strategize a way to get a better score, or to beat the game. But, that’s not the goal here. “Mercer envisions that the applicant will play the game before they even become a candidate. It can be done on online and even on a mobile device. It’s a two to three minute game that’s designed to be played once and only once. You don’t want applicants to practice because you want their natural traits and tendencies to come out. In fact, they won’t be able to figure out how to play the game, just give their natural answers. They won’t know the top performers’ answers, so they can’t game the game,” said Marder.
Increased diversity across industries
Because the Pymetrics’ approach is based solely on data and algorithms, it doesn’t look at a person’s gender, age or race. A candidate’s game results are matched to the company’s top performers’ results. Period. That means this form of predictive hiring helps increase diversity. “We screen in more diverse candidates, in some cases a 5-times improvement,” said Marder. “We feel the system is biased for those who’ve gone to best schools or have the best experience. Using predictive hiring tools create opportunities to level the playing field and to give opportunities to people who generally haven’t had them.”
Predictive hiring models like this one can also be applied to most jobs in many industries. “We think that there’s an interesting application for sales, retail sales, and call center sales - the sales DNA is easier to pinpoint, allowing the program to direct people to a sales career – but it crosses all industries. The model can be used in any role in any industry to tell us these are our top performers and tell us what is driving their performance,” said Marder.
Perhaps the only drawback for now using these kinds of predictive hiring models is that scale is needed to build a library of top performers by which job applicants can be compared to. For example, if you’re a small company, perhaps 10 to 50 total employees, it will be hard for you to use one of these predictive hiring models because you don’t have enough employees to provide the critical mass of top performers to compare new hires against. Or maybe you’re a new company and you haven’t identified top performers yet. But if I may borrow a phrase from an eagerly-anticipated movie that just came out: patience, young Jedi. Mercer is creating top-performer profiles that can be used in just about any setting. “Ultimately the size of the company doesn’t matter. We’ve created 40 different profiles that will be benchmarked,” said Marder. “Plus, the models get smarter and smarter the more people use them.”