article » There's An Algorithm That Reveals How Long Workers Will Stay At A Job

There's An Algorithm That Reveals How Long Workers Will Stay At A Job

November 19, 2012
4 min read

A common assumption among employers is that job-hoppers and the long-term unemployed make bad employees. The reasoning is that job-hoppers – defined as those who have held more than four full-time jobs during a five year period or more than four positions for less than six months – or those people who are currently unemployed, lack a commitment to long term job tenure or are at a skill deficit compared to peers with similar backgrounds but who are currently employed.

In today’s difficult economy, this common myth is so pervasive that job seekers fudge tenure at companies or shorten gaps on resumes to improve the perception of their competitiveness. Data from ADP and The Society of Human Resource Managers shows that 53 percent of candidates falsify their resumes. These same reports also show that 70 percent of college students said they would lie on a resume to get a job they want.

But for the five million individuals recorded as having been jobless for 27 weeks or longer by the U.S. Bureau of Labor Statistics in their August report, this fudging and shortening of gaps is no longer a viable option. These five million individuals must either lie to better position themselves to get hired, or accept that their unemployed status might be a permanent barrier to reentering the workforce.

According to a new study released by workforce intelligence company Evolv, these tactics should not be necessary. The study found that in the hourly workforce, prior history of unemployment has no bearing on performance or attrition. Likewise, evaluating applicants based on whether they had held multiple jobs for less than six months has no correlation to job tenure or performance.

Evolv utilized a robust sample of over 20,000 hourly employees and found that there was virtually no distinction in tenure between applicants that had held no jobs in the last five years and applicants that had held many jobs over the past five years. It then stands to reason that companies that continue to hire hourly talent based on these assumptions are unnecessarily exacerbating the problem of perpetual unemployment for hundreds of thousands of qualified job seekers.

This same study by Evolv showed that although job-hopping and a prior history of unemployment did not predict future job performance, other variables did. Employees who actively used one or two social networking sites on a weekly basis stayed an average of nine days longer than those who didn’t use social networks at all. Another indicator of attrition was familiarity with other employees at the company. The study showed that employees who knew three or more people working at the company were more likely to stay than those who knew none.

What does all of this mean for employers? Instead of using out of date methods based on gut feel or assumptions, employers can now leverage new technology and information to not only manage but also significantly improve the overall effectiveness of their workforce.

Utilize better technology

Technology is key to smarter workforce management and ultimately to improving the effectiveness and profitability of the workforce. According to The Wall Street Journal, Starbucks attracted 7.6 million job applicants for 65,000 openings in 2011. Though the applicant to job opening ratio is relatively high for companies like Starbucks, these record rates of applications are fairly common and demonstrate a clear need for solutions to help select and manage the workforce.

Unlike a few years ago when workforce technology basically meant a transparent, cheesy pre-hire test, there are now sophisticated, comprehensive solutions that take a holistic approach to workforce intelligence. These technologies focus not only on selection of new hires but incorporate and take into account the ongoing relationship of employers to employees from initial hire to eventual separation.

Evolv, for example, cites that for a recent client, this approach to workforce management meant a 25 to 35 percent reduction in attrition and a 20 to 30 percent improvement in performance across their entire workforce.

Leverage Data to Drive Decisions Pre and Post Hire

Big data makes it possible for scientists to decode the human genome and understand previous Rest assured, hiring managers — you can now identify candidates prone to job hopping.

Evolv — a data provider — has come up with a formula that predicts how long a candidate will likely stay with a specific company.

The algorithm is based on employment history, performance data and social media usage combined with certain econometrics such as gas prices and nationwide unemployment rates.

“These models will predict your likelihood of separating from your employer based on everything we know about you and the position,” Michael Housman, managing director of analytics at Evolv, told Christina Farr at VentureBeat.

Farr said that there are two reasons why people leave their jobs — it doesn't fit their skills-set or it's not the right cultural fit.

With the adequate data, Evolv said it can place people in jobs where they'll likely stay the longest.

Here's what the research has concluded so far:

An individual employee’s success is directly affected by their manager's performance.

High performance employees are those with exactly two social media accounts — any more or less will make employees "less likely to succeed."

The most loyal employees are those with personality traits like reliability and curiosity.

Evolv is currently collaborating with The Wharton School of the University of Pennsylvania for further research on employee retention.