Skip to main content
Next response time (NRT) is how long a customer waits for each follow-up reply after the first one. Where FRT covers the opening exchange, NRT covers everything after that.

How it’s calculated

NRT measures every wait after the first reply: from a customer message to the next reply. Several customer messages in a row count as one wait, starting at the first of them. A thread with 3 customer follow-ups contributes 3 measurements, each counted as a separate data point.
  • Waits answered by a machine user, such as Ari or an API integration, are left out.
  • The window covers threads created in the selected period, not threads whose replies were sent in that period. A reply sent today on a thread created 3 months ago counts toward the period 3 months ago.
  • Median and 90th percentile are computed across all individual gap measurements, not per-thread averages.

Statistical views

Median: the midpoint gap across all follow-up reply pairs in the selected period. 90th percentile: 90% of follow-up gaps were shorter than this. Use it to identify the slowest-moving conversations.

Team member view

In Team reporting, NRT covers the waits that team member ended by sending the reply.

SLA tracking

NRT has its own SLA target, separate from FRT. A thread breaches its NRT SLA if a human follow-up reply does not arrive within the configured window. See SLA compliance for breach rate details.