Decline in Workday Habits May Begin Seven Weeks Prior to an Employee Departure
New data from the Prodoscore Research Council (PRC), an initiative created by employee visibility and productivity intelligence software leader Prodoscore, identifies several key indicators that help detect behavioral patterns characteristic of employees likely to leave an organization.
The lead researcher for the PRC, made up of HR, Future of Work and productivity experts, evaluated more than 2M data points generated by 3,000 – 5,000 employees over a 12-week period.
“Overall, I’m observing employees working fewer total hours week-over-week for up to two months prior to leaving an organization,” explained Adrian Reece, PRC Principal Statistical Consultant and Doctoral Student. “Not surprisingly, employees become more disengaged as they draw closer to their final week of employment.”
HR Technology News: Conduent Ranked Among Top Global Companies for Best Culture by Comparably
PRC’s analysis of the behavior of employees likely to leave, compared to those likely to stay, provides insight into workforce attrition. Prodoscore’s data gives leaders an opportunity to quickly identify red flags and immediately address employees’ concerns or make appropriate changes within the organization.
Here is what the data revealed:
Employees who are poised to leave demonstrate, on average, more volatility in their start and end times during the three weeks prior to leaving, when compared to employees who stay.
- Individuals who stayed realized a one-minute increase in the time they started work and a one-minute extension in the time they ended their day
- Individuals who left realized a 16-minute decrease in the time they started work and a 24-minute reduction in the time they ended their day
The workday starts later and ends earlier for those likely to attrit. Five-to-seven weeks prior to leaving there is a reduction in hours worked during the average workday for people who intend to quit, which becomes more pronounced closer to departure.
- Twenty-one days before an employee quits, people staying with the organization put in an average of 1 hour and 31 minutes more than the person(s) leaving
- Seven days prior to the employee’s last day, people staying worked an average of 2 hours and 7 minutes more than the person(s) leaving
The rate of change in an employee’s Prodoscore, email volume and calendar time are indicative of whether or not an employee is likely to attrit.
- When looking across a 12-week period, employees who leave registered a decline in their Prodoscore (down an average of 33.5%), calendar time (down an average of 33%), and email activity (down 51%). Those three key indicators provide 89% accuracy in predicting attrition or retention. The change in behavior begins to occur 35-to-49 days prior to an employee leaving the organization, with the most dramatic changes taking place approximately two-to-three weeks prior.
Interaction with coworkers decreases drastically for employees who leave.
- For those likely to attrit, there was a 52.5% decrease in voice and video activity and a 52.5% decrease in messaging/chat activity, suggesting that interaction with coworkers plummets prior to an employee leaving.
HR Technology News: e-JAM Provides Full-Time, Contract, and Freelance Talent Out of Jamaica
“Insight into potential attrition is a huge benefit to leadership. Being able to mitigate surprise resignations, or prepare for them, can save businesses thousands of dollars. The implications are far-reaching and will allow managers to make better informed decisions in general,” said Prodoscore CEO Sam Naficy.
“Technology is truly allowing us to see behaviors that go unseen,” added Reece. “Using data to assess how people actually show up to work differently offers a new vantage point to inform leaders on what patterns and performance fluctuations are normal behavior and when it’s best to intervene.”
HR Technology News: Butterfly Network Appoints Timothy Trodden to Management Team