Service-Desk SLA: why the response-time metric hides the real problem
Most Service-Desk contracts treat SLA as if it were a single metric: response time or resolution time within an agreed limit. It's an easy metric to measure and easy to put on a dashboard, but it measures how fast a ticket leaves the queue, not necessarily whether the user's problem was actually solved.
Gartner itself flags this distinction in its research on automated support: today's AI tools can deflect more than 45% of incoming requests, pulling the ticket out of the queue quickly, but only about 14% of them reach genuine self-service resolution, with no ticket reopened within 72 hours. The gap between those two numbers, 45% versus 14%, is exactly the space where a response-time SLA looks great on the monthly report while user satisfaction stays poor.
That happens because response time and effective resolution measure different things. A ticket can be "answered" in minutes with a generic fix that solves nothing, and the user reopens the same ticket, or opens a new one, days later. From the SLA report's point of view, both tickets count as two on-time interactions. From the user's point of view, it's the same unsolved problem twice.
The most common symptom of this gap is the reopen rate, a metric most Service-Desk contracts simply don't track, because looking only at first-response time is easier to report. When reopens aren't measured, the support vendor gets rewarded for closing tickets fast, not for resolving them well, which creates a structural incentive for shallow answers.
The root cause, in most cases, isn't a lack of technical skill on the analyst's part, it's a lack of access to the right fix at the right moment. An analyst who has already seen that symptom before, or who has access to a record of how a similar case was resolved, delivers a quality answer within the same time it would take to deliver a generic one. Speed isn't the problem, what fills that speed is what makes the difference between deflecting a ticket and solving a problem.
That's why a well-designed Service-Desk SLA should include, alongside response time, at least two other metrics: reopen rate within a defined window (72 hours is the standard Gartner itself uses to separate deflection from genuine resolution) and first-contact resolution rate. Without those two, an SLA contract measures how fast the queue empties, not how well problems get solved.
At Fiscalizei, every ticket handled through the Company Brain carries the history of the fix applied and its outcome, which lets us measure reopens and first resolution natively, not as a separate manual report. That changes the incentive: both the AI agent and the human analyst get measured on truly resolving the issue, not just on clearing the ticket out of the queue on time.