Skip to main content
The module monitors two deadlines: the first answer and the solution. Both are set in minutes in the SLA card settings. An empty field means no term, and it’s not the same as zero: zero as a value is not accepted. The decision period cannot be shorter than the first response period. The timeline is set to the ticket when it’s created, and then the panel shows each of them the deadline and the status: time, waiting or overdue.

Pauses

SLA pauses only in two states: “Expects the customer” and “Expects the internal action.” The accumulated pause is added to the deadline, so waiting for the client’s response does not burn the deadline. In line is not considered a pause, and this is a conscious decision: an understaffed queue is exactly the problem that the SLA should make visible, not hide.

Delays

The delay is calculated at the time of reading, not from a flag that someone has to put in time. A ticket burned at night is read in the morning burned. A background check once a minute additionally marks fresh delays of the first answer and writes the event in the history of the ticket, one per ticket. There’s no automatic delay, no reassignment, no escalation, and it’s up to the team to decide, and you can see delays in the queue through the “SLA Violated” filter and in the analytics.

Where does waiting go?

In addition to timing, the module separately saves three time meters: how much the ticket was waiting for the client, how long the operator waited (states “New” and “In line”), and how much he spent in internal expectations, this breakdown is visible in the analytics in the card “Where the wait goes” and helps distinguish “team is not in time” from “the client is gone”.
Start with soft deadlines and squeeze them on the actual percentiles from the Time to Answer and Decision card in Analytics, not vice versa.