Average handle time is the metric most helpdesks track and the one most easily gamed, because an agent under pressure to keep it low learns to close tickets fast rather than resolve them properly, and the same problem returns as a second ticket a week later. The metric improves while the actual experience of being supported gets worse.
First-contact resolution — the proportion of tickets closed without the requester having to come back — is harder to fake, because gaming it means actually solving the problem the first time. It also correlates far more closely with user satisfaction than handle time does, because what frustrates people isn't a five-minute call, it's the same issue happening three times before someone finally fixes it properly.
Improving it usually isn't about agent skill, it's about what the agent has access to at the point of contact. An agent who can see the requester's device history, recent tickets and account status resolves more on the first call than one working from the ticket description alone. Most of the investment that moves this number is in tooling and knowledge access, not in training.
The trap is measuring it per-agent and using it for individual performance reviews, because agents respond by declining to close ambiguous tickets rather than by genuinely resolving more of them — moving borderline cases to 'pending' rather than risking a reopened ticket against their name. Measure it at the team level, aimed at fixing the systemic gaps — the missing knowledge article, the tool nobody has access to — rather than at individual accountability.
