How to use predictive analytics to reduce employee turnover?

June 1, 2019

 Employee turnover costs, a lot. 

 

Take this example. Turnover at 19% annually (meaning that 19 out of 100 people leave each year) will cost an organization in the IT industry (where avg. revenue per employee is around $400,000) in the neighbourhood of $80,000 a year in direct replacement costs, and an astonishing $1,100,000 annually in lost revenue. Employee turnover eats money, especially when turnover includes some of your top performers. 

 

An employee leaving is very clear to account for in retrospect - for instance, Mark had no promotion in 4 years, his manager has one of the lower performance ratings and Mark's pay is well below industry benchmarks. Plus, he's increasingly been late to meetings and once there, seemed overly critical. Putting these things together into an intent to leave becomes obvious after he announces his departure - but the real business value comes from those practices that can help us identify the intent to leave prior to the person's exit from the organization. 

 

This is precisely what predictive analytics entails - it enables us to identify individual employee turnover risk, in other words, it allows us to target employees who are at a high risk of leaving the company with timely and proactive retention initiatives. How does predictive analytics do this?

 

 

 

Predictive analytics is a branch of data science used to make predictions about unknown future events. In the retention context, the analysis takes in a number of factors (such as time since last promotion, sick leave, relative pay, past reviews, managers' reviews and others) as variables likely to influence one's decision to leave. Many of these variables are historical data that organizations routinely collect on their employees. Once you decide on the data you'd like to include in the analysis, you can formulate statistical models and choose ones that make the most accurate predictions for the future. Turnover risk models, given the right variables and enough training data can achieve astonishing levels of accuracy (read 85-95 %). 

 

The analysis output revolves around 2 main sets of findings: (1) a designation of a "flight risk" probability for each individual employee in the sample and (2) identification of variables associated with high levels of turnover. If flight risk is high among the organization's top performers, taking proactive steps now can mitigate that risk and actually prevent talent/knowledge loss. The key to reacting is to find out the reasons for employees' intent to leave (through focus groups and/or stay interviews), so that you can devise targeted retention strategies. Sometimes, these can be as easy as adding a few days off, bumping up family-friendly benefits package or adding a few percentages on the bonus amount. 

 

Predicting turnover has a remarkable impact on the business. First, it helps steer a retention strategy (and budgets) in a very clear and targeted way. As a result, the company sees a significant decrease in voluntary turnover, lessening the financial stress on the company and managerial stress struggling to replace key performers on their teams. Finally, successful retention initiatives also take the stress out of the recruitment team and enable them to focus on adding, not just replacing valuable employees to your workforce. 

 

Most companies try to solve for the employee turnover retroactively - once the employee has announced his intent to leave, for example by conducting exit interviews. Exit interviews however have proven to be a rather unreliable source of real reasons why employees left, and have rarely resulted in targeted, future-focused retention strategies. Predictive analytics, especially when combined with a qualitative dimension (such as "stay interviews" for instance) give you the ability to look forward and proactively tackle these risks, as well as devise appropriate and targeted strategies for different risk groups.

 

People Analytics Hub has a data science team that uses predictive analytics to assess the flight risk for each individual employee, as well as more focused insights into the reasons why these people may want to leave and how to keep them. If you find this useful, please comment, share, like or contact us with any questions. If you'd like to be part of our HR Trailblazer case study series, apply at research(at)alathea.io for special treatment!

 

Share on Facebook
Share on Twitter
Please reload

Featured Posts

Design Thinking - recepti za inovacije

December 12, 2017

1/1
Please reload

Recent Posts
Please reload

Archive
Please reload

Search By Tags
Follow Us
  • Facebook Basic Square
  • Twitter Basic Square
  • Google+ Basic Square

© 2016 ALATHEA RESEARCH & STRATEGY