The same agent performs differently on different calls largely because of emotional contagion — a documented, mostly unconscious process where an agent’s emotional state shifts to mirror the caller’s, call after call, all day. A pleasant, patient caller tends to produce a pleasant, patient agent response. A hostile, agitated caller tends to produce a more irritated, defensive agent response — not because the agent decided to respond that way, but because the nervous system picks up and reflects the emotional tone it’s exposed to, often before the agent is consciously aware it’s happening.
The research behind why the same agent performs differently on different calls
A study observing real call center employees found that agents tend to unconsciously imitate the emotional state of the customer on the line — when a customer was bad-tempered, the agent became more easily irritated; when a customer was joyful and pleasant, the agent became more patient and warm in turn. The researchers confirmed this as a real instance of emotional contagion occurring specifically in voice-to-voice customer service interactions, not just a general theory borrowed from other settings.
This means call-to-call variability isn’t primarily a measure of an agent’s discipline or professionalism. It’s a measure of what each individual call is doing to the agent’s nervous system in real time, call after call, with very little recovery window between one interaction and the next.
Why this compounds across a shift
A single hostile call doesn’t just affect that one interaction. Because the agent’s emotional state has shifted, often without their full awareness, the next call — even if that caller is perfectly pleasant — can be colored by whatever state the previous call left behind. An agent who absorbed three difficult calls in a row is starting the fourth call from a different physiological baseline than the agent who just finished three pleasant ones, even though nothing about the agent’s training, motivation, or skill has changed between calls.
This is the call-level version of the same problem described on our page about what causes inconsistent agent performance day to day — except here the swing happens in minutes instead of across a shift, often multiple times an hour.
Why this gets missed in standard QA review
A QA evaluator usually reviews calls individually and in isolation, scoring each one against the same standard regardless of what call came immediately before it. That approach can’t see the contagion effect at all, because the thing that explains the dip — the emotional residue from the previous caller — isn’t present in the call being scored. The agent gets dinged for a flat or short response without any visibility into what they were carrying into that call. This same blind spot is described in more depth on our page about why scripts fail when regulation fails.
What this means operationally
If call-to-call variability is a real, expected, physiologically grounded pattern rather than a discipline issue, the more useful intervention isn’t telling agents to simply not let it affect them — that instruction ignores the actual mechanism. The more useful intervention is building genuine recovery capacity between calls, so an agent’s baseline resets faster after a difficult interaction instead of carrying forward into the next one. That’s the specific gap ORS™ is built to close, the same accumulation described in our page on operational dysregulation load.