In the war for talent, organizations cannot succeed without a proactive strategy for retaining key employees. A customer recently asked us: “Is our turnover problem contagious?” In other words, does losing a key employee increase the likelihood that their peers leave as well? They had an intuition that this might be true but lacked the data to be sure. And, if true, how can they stop it?
Network analysis is a useful tool with which to answer this question. By mapping out employee networks, using metadata from common digital tools, we can identify who people’s most important connections are at work. Combining this with data on employee turnover allows us to analyze the long term outcomes of employees following the departure of a close peer.
after an employee exits a team, their direct peers can be 2 to 3X more likely to exit as well
What have we found? Turnover can be highly contagious. We’ve found that exits are strongly clustered within peer networks. That is, after an employee exits a team, their direct peers can be 2 to 3X more likely to exit as well. The role of the individual and connectedness of the team also play a key role in this dynamic. The exit of a highly-connected individual (an influencer) is typically more impactful. This effect is even stronger in densely connected teams.
Image: Network analysis showing clusters of turnover
Why is turnover contagious?
To understand the mechanism of contagion, we interviewed customer employees and uncovered the following insights:
- Having someone near you leave is a key moment in your employee experience. It can impact how you perceive your team, manager and the organization as a whole. If they didn’t think it was worth staying, why should you? This effect is amplified when the person who left was perceived to be a highly competent top performer.
- Losing a close work friend can remove a strong incentive to remain in an organization. The peer left behind is often forced to reform their social group at work. Work spouses are an extreme example of this effect.
- The “long good-bye” of an unhappy employee leaves a toxic atmosphere. The impact on the morale of the team is evident, even months later.
- Correlation vs causation. Having someone leave is an indicator that there may be other deeper problems with the employee experience on a team. There may, for instance, be an inexperienced manager struggling to motivate and align the team. This common factor is likely to lead to further turnover within the same cluster of individuals if it goes ignored.
What can you do about it?
It’s clear from our analyses that the contagion effect of turnover is real. Given that, what can be done to prevent it from spreading? Here are a few of the suggestions we make to organizations we work with.
- Take action quickly. When any key employee leaves it’s critical that organizations have a process for reacting quickly. Waiting too long leaves the remaining team at serious risk.
- Conduct exit interviews. Having a great exit interview process is key. It ensures that organizations quickly figure out if there are other issues at play.
- Engage direct managers. Direct managers and other supporting leaders have a major influence over the team’s ability to bounce back after the loss of a teammate. It’s critical that they feel engaged and responsible for taking corrective action.
- Be transparent. It’s key to openly address any exit with the remaining members of a team. If there are deeper issues at play, discussing them and taking action is the best way to ensure they don’t recur.
- Proactively identify issues. We’ve worked with several organizations to proactively identify problems with employee experience. Taking action before small issues grow into larger problems can significantly decrease the risk of turnover in the first place.
The data is clear: employee turnover can be highly contagious. This is yet another reason why organizations must work to proactively retain talent by focusing on employee experience. And when turnover does happens, it’s critical that leaders react quickly to prevent further spread.