SLA policies and timers
The four SLA timers, how deadlines are set, pause and resume on customer waits, and the at-risk, critical, and breached stages.
A service level agreement (SLA) policy sets the response and resolution targets a conversation must meet. Hark tracks up to four timers per conversation and shows, at a glance, how close each one is to breaching.
The four timers
| Timer | What it measures |
|---|---|
| First response | Time from arrival to the team's first reply. |
| Next response | Time to each subsequent reply while the customer is waiting. |
| Resolution | Time from arrival to the conversation being resolved. |
| Queue wait | Time a live chat waits in the queue before an agent accepts. |
Each target is set in minutes, and any timer can be left untracked by leaving its target empty. A policy can be restricted to a single priority (for example, a tighter target for URGENT) or apply to all priorities.
How a policy is chosen
A conversation gets its SLA policy from the inbox default or from a routing rule that overrides it. Once set, Hark computes the due time for each tracked timer from the policy's targets and, if the policy is attached to an office-hours calendar, measures only business time.
Pause and resume
The clock should not run against you while you are waiting on the customer. When a conversation moves to waiting on the customer (or waiting on an internal action), Hark pauses the SLA clock. On resume, each active deadline is pushed forward by the business time lost during the pause, so a customer who took a day to reply does not instantly breach your target. Snoozing a conversation holds the clock the same way.
Breach stages
For each timer, Hark surfaces how much of the target has been consumed and flags three escalating stages:
- At risk: the deadline is approaching.
- Critical: the deadline is imminent.
- Breached: the deadline has passed.
The queue-wait timer is measured from when the chat entered the queue, not from when the conversation was first created, so a case that sat as a bot chat for hours before escalating is not shown as instantly critical.
Configuring SLA
SLA policies and office hours are configured in Settings and are an Admin and Owner task. Agents work within the targets; managers set them.