Most service desk dashboards lead with the same two numbers: First Call Resolution and Customer Satisfaction. Both matter, but by the time either one moves in the wrong direction, the underlying problem has usually been building for weeks. A few less-visible metrics tend to give earlier warning.
Reopen rate
A ticket marked resolved and then reopened within a few days is a stronger signal than a low CSAT score. It means the fix didn't actually fix anything — the agent (or the process) addressed a symptom, not the cause. Track reopen rate by category, not just in aggregate; a spike concentrated in one category points directly at a specific process or knowledge gap.
Time in each state, not just total resolution time
Total resolution time hides where the delay actually happens. A ticket that sits "Assigned" for two days before anyone picks it up looks identical, in a total-time metric, to one that was actively worked the whole time but hit a genuinely hard problem. Breaking resolution time down by state — New, Assigned, In Progress, Pending — usually reveals that most delay lives in handoffs and queues, not in actual troubleshooting.
Knowledge article usage on resolved tickets
If agents are resolving a high volume of similar tickets without ever linking a knowledge article, that's either a knowledge base gap or a discoverability problem. Both are fixable, but only if you're tracking it. This metric quietly predicts future FCR problems before FCR itself moves.
Ticket volume by hour, against staffing
Comparing incoming ticket volume by hour against actual agent coverage by hour sounds obvious, but it's rarely built into standard dashboards. A team can hit every SLA on average while still having a rough 9-11am window every single day, because volume patterns and shift patterns were never actually compared side by side.
Why these matter more in ServiceNow specifically
Because ServiceNow captures granular state-change timestamps automatically, all four of these metrics are available without any extra instrumentation — they just aren't surfaced in the default dashboards most instances ship with. Building a small custom Performance Analytics dashboard around these four numbers usually takes less effort than people expect, and tends to surface problems 2-3 weeks before they show up in FCR or CSAT.
Start with one, not all four
Trying to fix all four at once usually stalls. Pick whichever metric is currently the biggest unknown for your team, build visibility into it first, and let that one change how the next quarter's process conversations actually get resourced.