In their classic 1982 song “Should I Stay or Should I Go,” the Clash asked a question that on any given day, some employees in firms of all sizes are asking themselves. The answer to that question can have a huge impact on a company’s bottom line, and being able to predict who’s likely to leave would be of clear benefit. Consider the value of helping your company prepare for (and possibly mitigate) the vast and numerous costs of employee turnover, financial and otherwise.
In a recent Wall Street Journal article, Rachel Emma Silverman and Nikki Waller talk about how Wal-Mart, Credit Suisse and other firms are crunching data to see which workers are likely to leave (“The Algorithm That Tells the Boss Who Might Quit”). The authors discuss how employee turnover has become a bigger and more costly worry as labor markets tighten, and how firms like Wal-Mart, Credit Suisse and Box Inc. are turning to predictive analytics as an early warning system that can allow managers to take action before an employee might choose to leave.
Now there’s no magic data point, nor one “killer model” that works for all companies. And that makes sense if you think about it; predictive models are empirically derived from unique, historical data sets. Those data sets tell us about a particular population of employees in a particular company environment. What might apply for Wal-Mart is not necessarily the same set of factors that matter at Credit Suisse or Box Inc. But any company can find what works for their specific situation using the same approach. The basic strategy is to apply the right algorithms to their data and create a predictive model that can improve early detection of imminent employee departure and create opportunities for possible intervention.
It’s not just large companies that can use this strategy; small and mid-sized businesses can do it, too. There are many firms (three are mentioned in the WSJ article) that can provide these “Workforce Analytics” services with the right data. The trick is pulling together all your employee-related data in one place, then finding which specific “predictor variables” in your data for your company are the strongest indicators that someone on your team is getting ready to leave. With so much data to look at, however, you can’t necessarily expect someone on your team poring over spreadsheets with thousands of rows and hundreds of columns to accurately pinpoint the strongest predictors. Using Workforce Analytics solves this problem with finesse, because it’s “the algorithm” that sorts through all those data points and finds the correct answer for you, without guesswork or exhaustive effort.
What you do with the information provided by Workforce Analytics is still up to you, and it’s still your call on whether or how to intervene, for a given employee predicted to be at-risk of leaving. But wouldn’t you rather be making decisions grounded in data, instead of guessing? Or worse, having to react in a costly way that impairs your business’s ability to stay productive?
To learn more about how Workforce Analytics and predictive models can help you detect when your employees are humming “Should I Stay or Should I Go” to themselves, contact me via email at firstname.lastname@example.org.