Support software loves to give you dashboards. Charts, gauges, trend lines, and a dozen numbers that all sound important. For a large support organization, tracking all of them makes sense. For a small team, most of that dashboard is noise that makes you feel busy without making you better.
This guide cuts it down to the six metrics that actually change how well you serve customers, what a good number looks like for each, and a few popular metrics you can stop worrying about.
Why fewer metrics is better for small teams
A metric is only useful if it changes a decision. If you look at a number every week and never act on it, it is not a metric, it is decoration. Large teams can afford analysts to mine subtle signals. A small team needs a short list of numbers where a bad reading points to an obvious next action.
So the test for every metric below is simple: when this number goes the wrong way, do you know what to do about it?
1. First Reply Time (FRT)
This is the single most important support metric for a small business, because it is the one customers feel most directly. It measures how long a customer waits between sending a request and getting your first meaningful response.
What good looks like: under 4 business hours for email, under 1 hour for social and messaging, under 2 minutes for live chat.
When it goes wrong: if FRT climbs, the fix is almost always about workflow, not effort. Centralize your channels so nothing hides in an unwatched inbox, and let AI draft answers to routine questions so a person is not the bottleneck. We wrote a whole piece on why first reply time matters if you want to go deeper.
2. Resolution Rate
Speed without resolution is just a fast way to disappoint people. Resolution rate measures the share of conversations that actually reach a satisfactory ending, ideally without the customer having to write back a second time.
What good looks like: most teams should aim for 80 percent or more of conversations resolved without escalation or repeated back-and-forth.
When it goes wrong: a falling resolution rate usually means your answers are fast but shallow. Look at what is being reopened, and improve the source material behind those answers (your knowledge base, your macros, or your AI grounding content).
3. First Contact Resolution (FCR)
A close cousin of resolution rate, FCR measures how often a customer gets a complete answer on the very first reply, with no follow-up needed. It is the difference between "we responded" and "we solved it".
What good looks like: 70 to 75 percent is a strong target for most small businesses.
When it goes wrong: low FCR often means your first replies are asking for information you could have anticipated, or answering only part of a multi-part question. Train yourself (and your AI drafts) to answer the question behind the question.
4. Customer Satisfaction (CSAT)
CSAT is the direct voice of the customer: a short rating right after a conversation closes, usually a thumbs up or down or a 1 to 5 scale. It is the closest thing to a ground-truth measure of whether your support is any good.
What good looks like: 90 percent positive (or an average above 4.5 out of 5) is excellent; anything below 80 percent deserves attention.
When it goes wrong: do not just watch the score, read the low ratings. Five one-star ratings with comments will teach you more than a thousand happy clicks. Patterns in the complaints are your roadmap.
5. Ticket Volume by Category
This one is not about a single number, it is about the shape of your inbox. When you tag or categorize incoming requests, you learn what customers actually struggle with. The top three categories usually reveal a product problem, a documentation gap, or an automation opportunity.
What good looks like: there is no target here. The value is in the pattern. If 30 percent of your tickets are "how do I reset my password", that is not a support problem, it is a product or content problem waiting to be fixed once.
When it goes wrong: if one category balloons, treat it as a signal to fix the root cause, not to answer faster. The best support ticket is the one that never needed to be sent.
6. Backlog (Open Conversations Over Time)
Backlog is the number of unresolved conversations sitting in your queue. It is the early-warning light on your support dashboard. A stable backlog means you are keeping pace with demand. A growing one means you are falling behind before customers have even complained about it.
What good looks like: flat or trending down. The absolute number matters less than the direction.
When it goes wrong: a steadily rising backlog means volume has outgrown your current process. That is the moment to automate a category you have been handling by hand, before the backlog turns into a wave of angry follow-ups.
The metrics you can safely ignore
Not every number deserves your attention. A few common ones do more harm than good for small teams:
| Vanity metric | Why to ignore it |
|---|---|
| Total tickets handled | Rewards busywork, not outcomes. Handling more tickets is not a goal. |
| Average handle time | Pushes agents to rush and close conversations prematurely. Speed to resolution matters, raw handling time does not. |
| Number of replies per ticket | Fewer replies can mean efficient, or it can mean you closed too early. Meaningless on its own. |
| Agent leaderboards | In a small team, ranking colleagues against each other damages morale far more than it improves service. |
How to actually use these six
Pick a single day each week and look at exactly these numbers, in this order: is my first reply time healthy, are conversations getting resolved, are customers happy, and is my backlog under control. If all four are green, do not touch anything. If one is red, the metric itself points to the fix.
The goal of measuring support is not a prettier dashboard. It is to notice a problem while it is still small, and to know what to do about it. Six numbers, checked once a week, will do that better for a small team than fifty numbers checked never.