The Impact of Tenure on New Employment Prospects
Reducing attrition is one of the greatest challenges for industries such as fast casual dining, retail, and contact centers. Annual attrition rates of more than 100 percent are not uncommon in these fields, and it is estimated that hiring and training hourly employees can cost employers up to $5,000 for each new hire. Conventional wisdom suggests that weeding out job-hoppers and ushering in applicants with more stable employment histories can help to cushion this turnover blow.
But a new study by Evolv presents conclusive evidence that employment history has no bearing on how long an employee will remain in her new position. According to the resulting whitepaper—Does Previous Work History Predict Future Employment Outcomes?—the study analyzed job applicant data and employment outcomes from a sample of 20,000 call center agents.
The tenure of employees who had held four or more jobs in the past five years showed no difference when compared with those who had held one to three jobs in the past five years. Even the segment of "perpetually unemployed" applicants with no recent job experience frequently racked up the same tenure as the more "stable" employees.
"Fundamentally, what we've found in this study is that an employee's tenure in any given position is more a function of fit than his previous job history," says Michael Housman, managing director of analytics at Evolv. "Research has shown that other screening methods that give a more nuanced understanding of the applicant, as well as his personality, aptitudes, work style and technical skills, have far greater predictive value than overall work experience."
Unfortunately, prejudice against job-hoppers still exists. "Recruiters often have dozens of candidates for a single position, and they are looking for ways to narrow down the candidates," says Housman. "These screening techniques become a way to shorten the list, even though employers may be eliminating solid candidates by using bad metrics."
For the full-length version of the article, please click here.