Share By Sachin Shenolikar
With resources scarce and job openings still at a premium, getting the green light to add staff is reason to celebrate. But after a fist-pump and a yesss, reality starts to seep in: That coveted position must be filled quickly with a person who can excel from Day One and stay with the team for a long time.
The first step, assessing a job candidate’s skills, is pretty straightforward — after all, there is no shortage of background information on the Internet, and any number of tests can be administered to gauge expertise. The tricky part, however, is the other crucial factor that can determine whether a new hire is a stud or a dud: Fit. How do his or her personality and style work in the company’s culture?
That’s a question hiring managers have been trying to answer for years. Some have preferred to go off hunches developed during the interview process. Others have administered questionnaires to get a sense of an applicant’s work views and habits. Still, making a successful hire has been mostly a crapshoot. But that is changing, thanks to data mining.
Data mining, or predictive analytics, has been hyped as a game-changer in finance, healthcare and marketing. You can add human resources to that list.
It’s a fact that organizations are becoming vast and complex, with employees working remotely all around the globe and ages ranging from the Baby Boomer generation to Gen Y. It’s a major challenge for companies to find the most suitable talent from such a diverse pool of candidates. That’s why they are investing resources to crack the code of big data.
“[Companies] are more concerned that [candidates] won’t be a fit than having the laundry list of skills that were asked for in the job description,” says Tony Deblauwe, a Silicon Valley-based career strategist and founder of the site Work Babble. “That’s why you’re seeing bigger investment dollars to quantify it in a better way rather than sending [candidates] a personality test or going with their gut in an interview.”
The goal is to create an algorithm that can calculate the specific qualities of people who have been successful in roles like the one being filled — and match those qualities with applicants. To that end, current employee production would be tracked, with everything from internal chat logs to posts on social sites such as Facebook and LinkedIn potentially being included in the analysis.
“Is there some sort of formula that says, We’ve surveyed our culture, we’ve got sort of a culture map, and we’ve identified the gold star — the best practices of what and how,” says Deblauwe.
“There is more data available [about] the qualities of individuals that are successful in the company, so there is a desire to replicate that,” adds Lisa Rowan, vice president of human resources at IDC. “It’s about more insight into what you would call a ‘success profile’ than you used to have.”
It’s important to remember that the promise of data mining is not that it will find a perfect candidate with 100 percent accuracy. The appeal is that it can increase the chances of finding the right person and make the interviewing process more efficient. “Big data says over the thousands of profiles we could [input] on what that job looks like successfully, there will eventually be enough patterns to say, Look, we could probably increase your odds that this [type of] person will be a better fit with 15 or 20 percent more predictive accuracy,” says Deblauwe.
Deblauwe believes big data analysis could also help job candidates, who could potentially find out whether a company or position is a good fit before spending the time and energy to apply. “I think that’s a huge untapped market,” he says. “How do I enable you to be better at choosing roles that you would agree are a good fit?
For now, the bottom line is that predictive analytics are starting to help companies get on the fast track to targeting employees who fit their values and style. It’s the reason businesses big and small are trying to make sense of the mass of information that’s out there to help them in them make good hires.
“How do we take this data lake and create some organization to it? That’s where we are now,” says Deblauwe. “Once it gets sorted out, it will get to the point of true predictive analytics where it will spit out these pretty patterns and help make decisions. That’s why people are so excited about it.”