The response-time promise — what a service level agreement covers, how the timers work, and what a breach should trigger.
SLA. An SLA (service level agreement) in customer support is a commitment to respond to or resolve conversations within a stated time — for example, a first reply within four business hours. An SLA pairs each target with a measurement method and, in contractual settings, with defined consequences for missing it.
Support SLAs come in two strengths. Contractual SLAs are terms a customer signed — common in B2B, sometimes with credits attached to breaches. Internal SLAs (strictly, service level objectives) are the same targets adopted voluntarily as operating discipline. Either way, an SLA policy usually stacks several timers: first response, next response after each customer message, and total resolution — often with different targets per priority tier, so an outage gets a tighter clock than a billing question.
The timers respect the calendar you declare: outside your stated business hours the clock pauses, which is what makes a four-hour target survivable for a team that sleeps. Good tooling warns before a breach, while there is still time to act, rather than reporting it afterwards.
An SLA converts “we answer fast” from a mood into a testable promise — and, more usefully, into a queue discipline: when everything feels urgent, the conversation nearest its deadline is objectively next. Breach patterns are diagnostic too. Breaches clustered on one channel, one tier or one time of day point at a staffing or routing gap, not at effort.
The known failure mode is optimising the timer instead of the customer: quick contentless replies sent to reset a next-response clock. The dashboard turns green, the customer waits exactly as long. If SLA attainment rises while satisfaction doesn’t, look there first.
SLA attainment formula. SLA attainment = (conversations that met the target ÷ conversations covered by the target) × 100, measured inside the business hours the SLA names. Attainment of 95 on a four-hour first-response SLA means 19 of every 20 covered conversations got a first reply in time.
Set few targets and mean them: a first-response target per priority tier is a complete starting policy. Decide in advance what a breach triggers — an escalation, an alert, an owner — because an SLA without a breach path is decoration. And aim attainment below 100 on purpose; a target you never miss is a target set too loose to tell you anything.
Give your customers faster answers and your team their evenings back.
Desktop and mobile. Free for solo.