A financial services powerhouse faced a 60 percent turnover rate among its 30,000 call-center reps. predictive talent analyticsLeveraging predictive talent analytics, the organization reduced turnover by 30 percent and saved $5 million in the first year alone.
A major medical device company suffered from high attrition among its highly paid sales reps as poor performers were forced out the door. Applying predictive talent analytics to shave attrition by only 1 percent, the company realized $30 million in savings.
Numbers like these should make any business manager pay attention to predictive talent analytics. And that’s a good thing, because talent analytics too often gets siloed in the HR department. The fact is, talent analytics is much bigger than HR. And the fast-emerging practice of predictive talent analytics requires the attention of every line of business.
Out of the silo
HR has long applied analytics to metrics like attrition rates. But these efforts have been backward-looking: What have been the past patterns of employee attrition? Predictive talent analytics is much more useful, because it looks to the future: What will our attrition rate be this year, who are the people who will leave the company, and what can we do to reduce that turnover?
It has only been in the past few years that we’ve developed the technologies and the methodologies to aim this future-focused lens. Data scientists created the means to predict how voters will vote, or how patients will follow treatment protocols, or how borrowers will pay off debts. It wasn’t long before HR realized the same technologies and approaches could be applied to predicting how employees will behave around key metrics like attrition and performance.
But talent analytics is often limited to the HR department. And that’s a problem, because typically, the most important data that feeds into predictive talent analytics resides not in HR but in the line of business.
HR is responsible for activities like identifying talent, hiring employees, and tracking data such as when they were hired, what they’re paid, and when they had their last performance review. But whether an employee is an accountant, a sales rep, an engineer, or a manager, the data on their performance and outcomes usually is captured and stored not in HR but in the function or line of business.
It’s these kinds of metrics that enable predictive talent analytics to deliver the greatest value. If you can accurately forecast which employees will perform best in a sales role, say, or which will be poor sellers and should not be in sales, then you can deliver tremendous returns on your predictive analytics efforts.
HR will still be involved in predictive talent analytics, of course, and may even manage the program. But for predictive talent analytics to deliver on its promise, it can’t be siloed in HR. Instead, it needs the data, the support, and probably the funding of the lines of business.
First steps to the future
In getting started with predictive talent analytics, the first thing organizations should do is follow the money. Rather than going on fishing expeditions with your employee data to see what you turn up, start with metrics your executive team will see clear value in funding. That will likely be attrition rates or job performance for key roles where there’s currently high turnover or where a small improvement in performance would save a lot of money or drive a lot of revenue.
Then, a straightforward cost-benefits analysis should show you, based on a given investment in analytics technology or services, what the payoff will be for a 10 percent or 20 percent reduction in turnover rates, say, or a 5 percent or 10 percent improvement in sales performance. You’ll then need to watch your model run for a while, reporting on a monthly or quarterly or yearly basis, and communicating the actual results you’re achieving.
Many large organizations may already have the analytics software and the data scientists to do this on their own. Others may need to work with an outside consultant, at least initially, till they gain the expertise they need. The mistake you don’t want to make is to assume your HR people, for example, will all become data scientists. Rather, you want to bring in analytics experts, whether in HR or the line of business or in a function like marketing, who have the expertise to do predictive modeling.
If you need to make investments in technology or talent, those costs should be straightforward to justify, because you should be able to point to clear returns. And if your results are anything like the financial services and medical device companies I mentioned earlier, your lines of business will be lining up to do more predictive talent analytics.