Agent churn describes people leaving the support team. To act on it, define who left, over what period and why. A single percentage can hide very different events: voluntary departures, planned seasonal contracts ending, dismissals or internal promotions.
Choose a consistent calculation
One practical turnover measure is departures during the period divided by average headcount during the same period, multiplied by 100. State how you calculate average headcount. The average of the opening and closing counts is simple, while a more frequent average may better represent a team that changes substantially.
For example, three departures from an average team of twenty produce 15% turnover for that period. This is a hypothetical calculation, not a benchmark. Report the dates and actual counts alongside the percentage, especially for a small team where one departure can move the result sharply.
Separate the types of departure
Keep voluntary and involuntary departures distinct. Decide how transfers and planned seasonal endings are reported, then apply that definition consistently. Look at tenure, role, shift and manager only where the group is large enough to support a useful interpretation.
A cluster of departures during onboarding suggests different questions from departures after several years. Do not assume all turnover means the same problem, or that a correlation proves the cause.
Investigate the working conditions
Review exit feedback alongside conversations with current team members and operational evidence. Useful questions include:
- Did the work and schedule match the hiring description?
- Could agents find current policies and get decisions on exceptions?
- Was workload manageable during their actual shift?
- Did compensation and development expectations remain clear?
- Were performance targets consistent with the complexity of the queue?
- Did managers provide timely, actionable feedback?
Treat an explanation as a hypothesis to investigate, rather than choosing a fashionable retention program before understanding the problem.
Connect churn to service cost
Estimate recruiting, onboarding, supervised work and the time spent covering vacancies. Keep those costs separate from speculative lost revenue. Review support KPIs during staffing changes to see whether response, resolution and repeat contacts also changed.
Document useful knowledge before a person leaves and assign owners to unfinished cases. A well-maintained manual reduces dependence on memory, but it does not replace experienced judgment or a functioning escalation route.
Choose an action that matches the evidence
If new hires leave because the role differs from the advertisement, revise the hiring process. If agents cannot apply policy confidently, improve training and access to a supervisor. If promotion paths are unclear, define the responsibilities and evidence needed for progression.
Use first-contact resolution practice for common cases and review the exceptions that remain difficult. Automation may change the workload, but it does not automatically fix schedule, compensation or management issues.
Review the action, not just the headline rate
Name an owner, a review date and an early indicator for each change. For an onboarding improvement, that might be fewer repeated policy errors and clearer feedback from new hires before enough time has passed to assess departures.
Keep tracking the same turnover definition and avoid overinterpreting a short period with few people. The aim is to understand which working conditions your team can improve and whether those changes make the operation more stable.
