Where call center detractors come from.
Findings from 14,351 calls and chats evaluated by independent evaluators between October 2025 and October 2026.
Key finding: customers left unsatisfied on 2.1% of calls with no problems, 10.5% of calls with a serious problem and 32.5% of calls with a zero-tolerance violation. That is 5 times and 15 times the clean-call rate. More than half of all unsatisfied customers (55%) came from the 16% of calls that had a serious or zero-tolerance problem.
Key findings
- A zero-tolerance violation (major rudeness, mocking a customer, a data disclosure, a silent hold of two minutes or more, or a failed transfer) left 32.5% of customers unsatisfied, about 15 times the rate on clean calls.
- A serious problem, such as no empathy, a wrong answer or an unexplained resolution, left 10.5% unsatisfied, 5 times the clean-call rate.
- Missing a scripted step, like the courtesy closing, did not raise dissatisfaction: 1.4% versus 2.1% on clean calls.
- Failed or abandoned transfers were the single most damaging behavior: 59.6% of those customers left unsatisfied.
- 55% of all unsatisfied customers came from just 16% of calls.
Customers who left unsatisfied, by the worst problem on the call
Customers who left unsatisfied, by what the agent did
| Worst problem on the call | Left unsatisfied |
|---|---|
| No problems on the call | 2.1% |
| Call-handling misses only | 1.4% |
| A serious problem | 10.5% |
| A zero-tolerance violation | 32.5% |
| Behavior | Calls | Left unsatisfied |
|---|---|---|
| Failed or abandoned transfer | 47 | 59.6% |
| Didn’t confirm what the customer needed | 59 | 45.8% |
| Rudeness, major | 32 | 31.2% |
| Silent hold, 2+ minutes | 65 | 29.2% |
| Wrong policy information | 232 | 26.7% |
| Resolution not explained | 208 | 14.4% |
| No empathy | 1,038 | 13.0% |
| Missed the courtesy closing | 2,609 | 3.3% |
What this means for NPS
Unhappy callers are where detractors come from, and most of them are created by a small set of agent behaviors. Coaching that treats every miss the same spends most of its time on scripted steps that barely move satisfaction. Telemedic 500 counsels the behaviors that do, first and hardest.
Methodology
- Sample: 14,351 customer service calls and chats evaluated by independent evaluators from October 3, 2025 to October 7, 2026.
- Measure: at the end of each evaluated call, the evaluator recorded whether the customer left satisfied (yes, maybe or no). The rates above are the share recorded as “no”.
- Grouping: each call is grouped by the most severe problem flagged on it. Behaviors shown occurred on at least 30 calls.
- Limits: these figures describe evaluated calls, not survey responses, and they show association, not proof of cause. Results come from one evaluation program.
How to cite
Telemedic 500. “Where call center detractors come from: findings from 14,351 evaluated calls.” November 2, 2026. https://telemedic500.com/nps-research/
Last updated November 2, 2026.