HOW TO PREDICT TURNOVER ON YOUR SALES TEAM

IT’S NOT ENOUGH TO KNOW WHO YOUR STARS ARE. YOU NEED TO MAKE SURE THEY DON’T LEAVE.

COMPANIES worry about employee attrition in every department, but it’s especially costly in one function: sales. Estimates of annual turnover among the US salespeople run as high as 27 per cent – twice the rate in the overall labour force. In many industries, the average tenure is less than two years. While some attrition is desirable, such as when poor performers quit or are terminated, much of it isn’t – and every time a solid performer leaves, his or her company faces a number of direct and indirect costs. US firms spend $15 billion a year training salespeople and another $800 billion on incentives, and attrition reduces the return on those investments. Turnover also hurts sales: Positions may sit empty while companies recruit replacements, and the new employees must learn the ropes and rebuild client relationships. If managers could identify good salespeople who are at risk of quitting and take steps to retain them, their companies could realise substantial savings.

A new study by four marketing professors, led by V. Kumar, of Georgia State University, can help them do just that. The researchers examined more than two years’ worth of data from a Fortune 500 telecommunications company that sells consumer electronics and software services, and created a quantitative model – the first of its kind – to predict which salespeople were likely to quit. This

work builds on previous research by some of the same academics, who developed a method of estimating an individual salesperson’s future profitability (see “Who’s Your Most Valuable Salesperson?”

HBR, April 2015). Knowing who is most likely to drive profits is useful, of course, but the new research could add greatly to that value: By learning who is at high risk of leaving and why, sales leaders can address problems before star performers give notice.

The researchers studied data on 6,727 salespeople working in 1,058 stores, dividing it into two batches. One set of metrics dealt with how well each salesperson was doing; those numbers measured past performance (on the basis of revenue generated), customer satisfaction, and how often monthly quotas were met. The second set measured “peer effects”: the variation in performance among coworkers and the voluntary and involuntary attrition in each store. The study controlled for geography, store size, and demographics.

The researchers expected that salespeople with high ratings in historical performance and customer satisfaction would be less likely than average and low performers to quit, because the good marks would increase their sense of job security, their incentive payments, and their feeling that they controlled their ab ability to succeed – and that pr proved to be the case. When it came to quota attainment, ho however, the study showed an inverted-U-shaped di distribution: Here, too, hi high-performing salespeople w were less likely than av average performers to quit (managers did a good job keeping their stars happy), but so were low performers (their poor showing limited their opportunities at other firms). “It is the ‘middling’ salespersons who [are] likely [to] turn over,” the researchers write. Though those employees aren’t “A” players, the loss of them still hurts their firms, because they often constitute a large and profitable part of the sales force.

The biggest surprise concerned peer effects, which turned out to be the strongest predictor of quitting. The researchers theorize that in companies without much variation in performance, people are less likely to feel challenged and may have little incentive to work harder or smarter; they’re apt to leave instead. In settings with high voluntary turnover, employees often lose faith in the company’s strategic direction (because they see others jumping ship), and they tend to be more aware of outside job opportunities, partly because their networks include former colleagues who recently defected. And when there’s lots of involuntary turnover, employees may lack trust in managers, feel little job security, and move on. “An individual’s attitudes and intentions are heavily influenced by his or her environment,” the researchers write; the strength of the peer effects in the model suggests that turnover can be contagious.

This research is part of a broad trend of efforts to understand what events cause employees to seek greener pastures and what behaviours indicate that they may be doing so – issues of increasing relevance

in an era of tight labour markets and the growing use of analytics. For instance, research by the advisory firm CEB examined how events in employees’ personal lives, such as milestone birthdays and college reunions, spur them to take stock and to compare their careers with others’, often prompting them to job hunt (see “Why People Quit Their Jobs,” HBR, September 2016). And a study by researchers at Utah State and Arizona State identified 13 “prequitting” behaviours, likening them to poker tells; these include leaving work early, showing less focus or effort, and being reluctant to commit to long-term assignments.

One implication of the new study is that managers should pay careful attention to peer effects and consider conducting interventions in settings with little performance variation among employees and ones with rising levels of turnover. But Kumar says the larger message isn’t that firms should plug their data into the model predicting turnover at the telecom’s stores. Rather, it’s that big data can enable companies to identify variables that predict turnover in their own ranks. In the future, managers might routinely rely on datadriven dashboards labelling employees as being at high, moderate, or low risk of quitting. They could then decide which members of the high-risk group warrant interventions to help them stay put.