Ask ten support managers which metrics they track and you will get ten different dashboards, most of them crowded with numbers nobody acts on. The problem is rarely a lack of data. It is that the wrong metrics get top billing, the right ones get calculated inconsistently, and the whole set gets reported without anyone deciding what a bad number should trigger. This guide is the short version: the customer service metrics and KPIs that actually move a support operation, how to calculate each one, what a good number looks like in 2026, and how to keep the metric from quietly corrupting the behavior it is supposed to improve.
None of this requires a data team. It requires picking a handful of measures, defining them the same way every week, and tying each one to a decision. If a metric on your dashboard does not change what anyone does when it moves, take it off the dashboard.
The customer service metrics that matter most
The six metrics below are the core set. They cover the three things support leaders actually need to know: how customers feel, how fast the team responds, and how effectively it resolves. Every other metric is either a component of these or a segment of them (by channel, by agent, by customer tier). Start here, get these right, and add specialty metrics only when a specific question demands one.
| Metric | What it measures | 2026 benchmark |
|---|---|---|
| CSAT | How satisfied customers are after an interaction | 85% or higher |
| First response time (FRT) | How long a customer waits for the first reply | Email under 24h, chat under 90s, phone under 3 min |
| First contact resolution (FCR) | Share of issues solved in one interaction | 60% to 80% |
| Average handle time (AHT) | Average time to work a single contact | Track the trend, not an absolute |
| Resolution rate | Share of tickets actually resolved | 90% or higher over a period |
| Customer effort score (CES) | How hard it was for the customer to get help | Lower is better; benchmark against your own baseline |
Customer satisfaction score (CSAT)
CSAT measures how satisfied a customer is with a specific interaction, usually captured by a one-question survey sent right after a ticket closes: "How satisfied were you with the support you received?" It is the most direct read on how your service feels from the outside, which is why most teams treat it as the headline number.
How to calculate it: CSAT = (number of satisfied responses / total responses) × 100. On a five-point scale, "satisfied" usually means the top two boxes (4 and 5). A CSAT of 85% or higher is a common target for a healthy support team, though the right bar depends on your product and audience. The trap with CSAT is survey bias: response rates are low and skew toward the very happy and the very angry, so read it as a trend across hundreds of responses, not a verdict on any single ticket.
First response time (FRT)
First response time is how long a customer waits between reaching out and getting a first human reply. It is the metric customers feel most immediately, because silence after sending a message is the fastest way to make someone feel ignored. FRT is a measure of responsiveness, not resolution, and the two are easy to confuse.
How to calculate it: FRT = total time to first response across tickets / number of tickets, measured per channel because expectations differ wildly. In 2026, reasonable targets are under 24 hours for email, under 90 seconds for live chat, and under 3 minutes for phone. The honest version of this metric excludes automated acknowledgments; an auto-reply that says "we got your message" is not a first response, and counting it is how teams fool themselves. If you commit to response targets formally, they belong in a written customer service SLA so both sides know what "on time" means.
First contact resolution (FCR)
First contact resolution is the percentage of issues solved in a single interaction, with no follow-up, no reopened ticket, and no second email. It is arguably the most valuable metric on this list because it correlates with almost everything else that matters: research has found a roughly 1:1 relationship between FCR and CSAT, meaning each one-point gain in FCR tends to lift satisfaction by about a point.
How to calculate it: FCR = (cases resolved on first contact / total cases) × 100. A benchmark of 60% to 80% is typical for a well-run team. The measurement is harder than it looks: you have to define what "first contact" and "resolved" mean and apply them consistently. A ticket the customer reopens two days later was not resolved on first contact, even if the agent thought it was, so tie FCR to whether the customer came back, not to the agent's self-report.
Average handle time (AHT)
Average handle time is the average time an agent spends working a single contact, including talk or writing time and any after-contact wrap-up. AHT is an efficiency metric, and it is the one most likely to be misused. A low AHT can mean an efficient team, or it can mean agents are rushing customers off the line to hit a number, which tanks FCR and CSAT at the same time.
How to calculate it: AHT = total handling time / number of contacts handled. There is no universal "good" AHT because a password reset and a billing dispute are not comparable. Track AHT as a trend within similar contact types, and never reward a falling AHT on its own. It is only good news when FCR and CSAT hold steady or improve alongside it. Read in isolation, AHT is the classic vanity metric that makes a support floor worse.
Resolution rate and customer effort score
Resolution rate is the share of tickets that actually get resolved over a period, calculated as (resolved tickets / total tickets) × 100. A healthy team resolves 90% or more of what comes in; a rate well below that usually means tickets are falling through the cracks rather than being genuinely hard. It is a good backstop metric because it catches the queue quietly filling with abandoned work.
Customer effort score (CES) asks the customer how much effort it took to get their issue handled, usually on a scale from "very easy" to "very difficult." It matters because effort predicts loyalty better than delight does: customers rarely reward you for a smooth experience, but they punish you for a painful one. A high-effort resolution, where the customer had to explain themselves three times and chase two follow-ups, damages retention even when the issue technically got fixed. Track CES against your own baseline and watch the direction it moves.
How to actually use these metrics
Collecting the numbers is the easy part. The discipline is turning each one into a decision, and there are a few rules that separate a useful measurement program from a wall of charts nobody reads.
Pair every efficiency metric with a quality metric. AHT and FRT can always be gamed by cutting corners, so never look at them alone. Read AHT next to FCR and CSAT; read FRT next to resolution rate. A speed number that improves while quality falls is a warning, not a win.
Segment before you conclude. An average hides the story. Break every metric down by channel, by agent, and by customer tier before deciding what it means. A blended CSAT of 88% can mask a chat channel sitting at 70%, and a team-wide FCR looks fine until one queue is dragging it.
Measure consistently or do not measure at all. The single biggest source of bad support data is changing definitions. If "resolved" means one thing in January and another in March, your trend line is fiction. Write the formulas down, apply them the same way every week, and let your customer service ticketing system compute them automatically rather than reconstructing numbers by hand, which is where inconsistency creeps in.
Set a threshold that triggers action. For each metric, decide in advance what number is bad enough to do something about, and what that something is. FCR under 60% triggers a review of which issue types keep coming back; CSAT under 80% on a channel triggers a look at staffing or training there. A metric with no trigger is decoration.
How many metrics should a support team track?
Fewer than you think. The six above are enough for most teams to run a support operation well, and a manager can hold all six in their head and reason about how they interact. Adding a seventh or eighth is fine when a specific question demands it, such as backlog age when tickets are piling up, or cost per contact when you need to justify the budget. What you want to avoid is the twenty-metric dashboard where everything is measured and nothing is managed. Track the small set that maps to decisions, report it consistently, and add a metric only when you can name the decision it will drive.
Frequently asked questions about customer service metrics
What are the most important customer service metrics? The most important customer service metrics are CSAT, first response time, and first contact resolution. CSAT tells you how customers feel, first response time tells you how fast the team reacts, and first contact resolution tells you how effectively it solves problems in one interaction. Those three cover satisfaction, speed, and effectiveness, which is nearly everything a support leader needs to judge performance.
What is the difference between a metric and a KPI? A metric is any number you can measure, while a KPI is a metric you have chosen as a key indicator of success and tied to a target. Every KPI is a metric, but not every metric is a KPI. First response time is a metric on everyone's dashboard; it becomes a KPI when your team commits to a specific target for it and holds itself accountable to that number.
How do you measure customer service quality? Measure customer service quality with a mix of outcome metrics and perception metrics. CSAT and customer effort score capture how the customer felt, first contact resolution and resolution rate capture whether the issue was actually solved, and internal quality assurance scoring captures whether agents followed process. No single number covers quality, so read satisfaction, resolution, and QA together rather than relying on one.
What is a good CSAT score? A good CSAT score is generally 85% or higher, though the right target depends on your industry and customer base. What matters more than the absolute number is the trend and the segments beneath it: a stable 85% is healthier than a volatile 90%, and a strong average can still hide a weak channel. Read CSAT across hundreds of responses and broken down by channel before deciding whether it is good.
Metrics are only worth collecting if the operation underneath them is built to be measured, which is the wider point of treating customer experience operations as real work rather than an afterthought. For how the back office decides retention, see why CX is won in the back office, and for the operational practices that produce good numbers in the first place, see how to run a shared inbox and build a customer service knowledge base.