A helpdesk generates numbers the way a server generates heat - constantly, and mostly ignored. Five reports, read together for fifteen minutes a week, will tell you more truth about your support than any dashboard wall.
Jira Service Management ships with a set of built-in reports per project, and most small teams open exactly one of them - usually when a manager asks an awkward question at the worst moment. Used weekly instead of defensively, the same reports become the team's planning tool: they show whether demand is growing, where it concentrates, whether promises are kept, and whether customers leave happier than they arrived. This page names the five that earn their keep, what decision each one feeds, and - with equal honesty - the popular numbers that flatter the team while saying nothing.
Two lines on one graph: tickets arriving and tickets finishing. When the created line lives above the resolved line for weeks, the backlog is compounding no matter how busy everyone feels - the desk needs capacity, demand reduction, or both. When resolved crosses above after a push, you can see the recovery and say so with evidence.
This is the report to take to any conversation about hiring or scope. 'We feel swamped' is an opinion a manager can wave away; six weeks of arrivals outrunning completions, with the gap priced in tickets per week, is a staffing argument nobody has to raise their voice for.
The workload report shows open and resolved work per agent. Read it for balance: one agent holding triple the open tickets is either your hero or your bottleneck, and both are management problems - the first burns out, the second blocks a queue.
Read it with its honest limitation in the same breath: tickets are not equal units. One agent's ten password and access jobs may be lighter than another's three server incidents. Workload data starts a conversation about who carries what kind of work; it never finishes one. The moment this report becomes a public ranking, agents will optimise for ticket counts, and your data quality will quietly rot.
This report splits your kept promises from your broken ones, by goal. A healthy small desk keeps the large majority; the value is in the breach minority - which goals, which request types, which hours of the week. Breaches that cluster on one request type usually indict a process (an approval that always waits a day), not a person.
Two honest cautions. First, SLA percentages only mean something after your working-hours calendar is set correctly - otherwise you are measuring weekends. Second, a rising "met" rate achieved by quietly loosening goals will show up eventually in the satisfaction report, which is why these five are read together.
The request-type breakdown is the most strategically useful report a small team owns. It is the office telling you, in counted form, what it needs: if access requests spike after every new joiner batch, your starter process is broken upstream; if one application generates a third of problem tickets, that is a training, licensing or replacement conversation with data attached.
Review it monthly and act on exactly one theme at a time - fix the form, write the knowledge article, raise the upstream issue. A helpdesk that reduces next month's ticket count is doing a different, better job than one that merely survives this month's, and this report is how you prove the difference.
Resolved tickets can carry a simple star rating with an optional comment. Response counts will be low - that is normal everywhere, not a local failure - and the comments, good and bad, are worth more than the average. Read every low score personally; a one-star with a sentence of explanation is a free consultation.
Hold the number loosely: satisfied users answer surveys at different rates than angry ones, and a tiny sample swings wildly week to week. Treat satisfaction as a smoke detector, not a speedometer - it tells you something is burning somewhere, and the comments tell you where.
Reports fail as wallpaper and succeed as ritual. Fix a quarter hour - Monday morning works - where the team lead opens these five in order and asks three questions: Is the pile growing? Are promises kept? What one theme do we attack this week? Notes go in a running document, one line per week, so trends have a memory longer than anyone's mood.
Skip the vanity shelf: total tickets ever, average handle time to the second, agent activity heatmaps. They exist, they look managerial, and none of them has ever fixed a queue. The five above each end in a decision - that is the whole test.
See every screen of the Jira helpdesk setup live, from the empty project to the first real ticket.
Watch Full VideoFive: created versus resolved for backlog direction, workload for balance, SLA met versus breached for promise-keeping, requests by type for demand patterns, and customer satisfaction for experience. Fifteen minutes with all five beats an hour with any one of them.
Because that is normal. Only a fraction of users rate a resolved ticket, and unhappy and delighted users respond at higher rates than the contented middle. Use the score as a warning light and read the comments - do not manage the team to the decimal point.
They show volumes, times and SLA results per agent, but raw counts mislead: ticket difficulty varies enormously, and ranking agents on counts teaches them to chase easy tickets. Use per-agent data for workload balance and coaching conversations, not public league tables.