Walk into any support org and you’ll find a dashboard glowing with numbers. Walk in a little further and you’ll find that most of those numbers don’t change any decisions. Measuring everything is the same as measuring nothing — it just costs more.
Here are the seven metrics worth building your operation around, and three that look impressive but quietly mislead you.
The 7 that matter
1. First Response Time (FRT)
How long a customer waits for any meaningful reply. It’s the single biggest driver of early satisfaction, because waiting in silence is what makes people angry — not the problem itself. Track it at the median and the 90th percentile; averages hide your worst experiences.
2. Full Resolution Time
How long until the issue is actually solved. A fast first response followed by three days of back-and-forth is not a win. Resolution time is the metric customers actually feel.
3. CSAT (Customer Satisfaction)
The post-interaction “how did we do?” score. Simple, direct, and predictive — but only if you act on the comments, not just the number. The 1-star reasons are a free roadmap for fixing your product and your process.
4. Re-open Rate
The percentage of “resolved” tickets that come back. This is the truth serum for your other metrics. A team can game resolution time by closing tickets early — re-open rate catches it instantly. Anything above ~10% means you’re closing things that aren’t solved.
5. Contact Rate (tickets per active customer)
Volume normalized by how many customers you have. Raw ticket count always rises as you grow, which tells you nothing. Contact rate tells you whether each customer needs more or less help over time. Falling contact rate is one of the best signals that your product and self-service are improving.
6. Automation / Deflection Rate (with quality attached)
The share of contacts resolved without a human — but only meaningful when paired with re-open rate and CSAT on those automated resolutions. A high deflection rate with rising re-opens is a fire, not a feature.
7. Agent Capacity & Backlog Trend
Not a customer metric, but it predicts every customer metric. If your backlog is growing faster than your team can clear it, response and resolution times are about to degrade no matter how good everyone is. Watch the trend, not the snapshot.
A good rule of thumb: every metric on your dashboard should be able to trigger a specific decision. If you can’t name the action a number would prompt, take it off the dashboard.
The 3 vanity metrics to ignore
Total ticket count (in isolation)
It only ever goes up as you grow, and “we handled more tickets” is not a goal. Use contact rate instead.
Average handle time, used as a target
Tracking how long agents spend per ticket is fine for capacity planning. Turning it into a target is poison — it pushes agents to rush, close early, and dodge hard tickets, which spikes your re-open rate. Optimize for resolution quality; let handle time fall as a result.
Number of help articles published
A 400-article help center that nobody can search is worse than 30 great articles. Volume of content is an activity metric, not an outcome. Measure article usage and deflection per article instead.
How to put this together
You don’t need a fancy BI stack to start. A weekly review of these seven, with one owner per metric and a short note on “what changed and why,” beats a real-time dashboard nobody reads.
The pattern that separates strong support orgs from struggling ones isn’t the tooling — it’s that they pick a small set of honest metrics, watch the trends, and actually change something when the numbers move.
If you’re rebuilding your reporting from scratch, it helps to use a platform that exposes resolution quality and automation metrics together rather than in separate silos. AItocha CX surfaces deflection rate alongside re-open and CSAT data, which makes it harder to fool yourself with a deflection number that isn’t actually solving anything.