Call Center Metrics: The 10 KPIs Small Teams Should Track (With Formulas and Benchmarks)

Call center metrics are the numbers that show whether callers reach a person quickly, get their problem solved and leave satisfied. For a small team, ten of them do almost all the work: first call resolution, average speed of answer, service level, call abandonment rate, average handle time, hold time, transfer rate, customer satisfaction, agent occupancy and call volume by hour.

You do not need an enterprise contact center to track them. Any phone system with call queues and reporting can produce most of these figures every day. This guide gives the exact formula for each one, a benchmark from a source you can check, what usually moves the number, and how to read the metrics together so improving one does not quietly break another.

What Are Call Center Metrics?

Call center metrics are measurements taken from your phone system and your customers that describe how inbound and outbound calls are handled. Some measure speed, such as how long callers wait. Some measure effort, such as how long each call takes. Others measure outcome, such as whether the issue was fixed on the first call and how the caller rated the experience.

The term is used loosely alongside call center KPIs. A metric is any number you can measure; a KPI is the handful of metrics you have decided to manage the team against, with a target attached. A five-person office might log twenty metrics but hold itself to three or four KPIs, reviewed every week.

Which Call Center KPIs Matter Most for a Small Team?

If you can only watch a few, start with the ones that tell you whether customers are being served at all, then add the ones that explain why. In our experience setting up queues for small businesses, this order works well:

• Call abandonment rate and average speed of answer: are callers reaching anyone?
• First call resolution: are problems getting fixed, or coming back as repeat calls?
• Customer satisfaction: does the caller agree the call went well?
• Average handle time and hold time: how much effort each call takes, and where it goes
• Call volume by hour: when you need more people on the phones

The rest of the list, transfer rate, service level and occupancy, becomes useful once you have a queue with several agents and want to fine-tune routing and schedules.

Call Center Metrics Formulas at a Glance

Use this table as a reference card. Every rate is a percentage, so multiply by 100. Times are in seconds unless your reports use minutes, and the calculation period should be the same for every metric you compare, usually a day or a week.

Metric	Formula	What it tells you
First call resolution (FCR)	Issues resolved on the first contact ÷ total issues	Whether problems are fixed without a callback
Average speed of answer (ASA)	Total wait time of answered calls ÷ calls answered	How long callers wait before a person picks up
Service level	Calls answered within the threshold ÷ calls answered	Share of callers answered quickly, such as within 20 seconds
Call abandonment rate	Calls abandoned in queue ÷ calls offered to the queue	Share of callers who hang up before reaching anyone
Average handle time (AHT)	(Talk time + hold time + after-call work) ÷ calls handled	Total effort per call, including wrap-up
Average hold time	Total hold time ÷ calls handled	How long agents park callers mid-conversation
Transfer rate	Calls transferred ÷ calls handled	Whether calls reach the right person the first time
Customer satisfaction (CSAT)	Ratings of 4 or 5 ÷ total survey responses	How callers rate the experience on a 1–5 scale
Agent occupancy	Time handling calls ÷ time logged in and available	How busy agents are while on shift
Call volume by interval	Calls offered per 30- or 60-minute interval	When demand peaks, for scheduling

How to Calculate and Improve Each Metric

Formulas are the easy part. The judgement is in what you include, which calls you exclude and what a reasonable target looks like for your size. Here is each metric with the details that change the answer.

First Call Resolution

First call resolution is the share of customer issues solved on the first contact, with no follow-up call, email or visit. It is the metric most closely tied to how customers feel about the call. SQM Group's FCR benchmark of more than 500 North American call centers put the 2021 average at 71%, with centers ranging from 44% to 92%, and SQM reports that every 1% gain in FCR comes with roughly a 1% gain in customer satisfaction.

Small teams rarely run post-call surveys, so measure FCR with a repeat-call check instead: count callers whose number calls back about the same issue within a set window, often seven days, and treat those first calls as unresolved. Agents can also tag a call as resolved or follow-up needed during wrap-up. Either way, write the rule down, because a two-day window and a 30-day window produce very different FCR numbers.

Average Speed of Answer

Average speed of answer is the average time a caller waits in queue before an agent picks up. Only answered calls count, which is why ASA should always be read next to the abandonment rate: a queue can show a good ASA simply because the people who would have waited longest hung up first.

ASA is shaped mostly by staffing against demand. If it climbs at the same hours each day, the fix is usually a schedule change, not faster agents. Our call center staffing calculator shows how many agents you need per interval to hit a wait-time target.

Service Level

Service level is the percentage of calls answered within a time threshold you choose. The long-standing industry standard is 80/20: 80% of calls answered within 20 seconds. In how SQM calculates service level, the formula is calls answered within the threshold divided by total calls answered, and SQM notes that only 16% of call centers consistently meet the 80/20 standard.

The detail that matters is how you treat abandoned calls. Excluding every abandoned call flatters the number. A fairer method excludes only short abandons, callers who hang up within the first 5 to 10 seconds, often because they dialed the wrong number, and counts everyone who gave up after a real wait.

Call Abandonment Rate

Call abandonment rate is the share of callers who hang up after entering the queue but before reaching a person. The General Services Administration's IT Modernization Centers of Excellence suggests keeping abandoned calls under 3% as a goal for government contact centers, a useful ceiling for any team that wants callers to feel answered.

Abandonment usually rises with wait time, so it responds to the same staffing fixes as ASA. Routing helps too: a queue that holds callers in order and announces their position keeps more of them on the line than a ring group that rings everyone and then drops to voicemail. Our guide to call queue vs ring group explains when each makes sense. A callback option, where the caller keeps their place without staying on hold, also cuts abandonment during peaks.

Average Handle Time

Average handle time is the total time an agent spends on a call, from the moment they answer to the moment they finish the related work. It adds talk time, hold time and after-call work such as notes, tickets or follow-up texts, then divides by calls handled. Leaving out wrap-up time is the most common mistake and makes AHT look shorter than the real workload.

There is no universal good AHT. A pizza order and a billing dispute should not share a target. Compare AHT by call type and by agent instead: a gap of several minutes between two agents handling the same calls is a training opportunity, while a team-wide rise often points to a new product, policy or system that callers need explained.

Average Hold Time and Transfer Rate

Hold time and transfers are both signs that the person who answered could not finish the call alone. Long holds usually mean an agent is searching for information or waiting on a colleague. A high transfer rate usually means the auto attendant menu or routing is sending callers to the wrong place first.

Review the calls behind these numbers before changing anything. Listening to a handful of held or transferred calls through call recording often reveals one missing answer, one confusing menu option or one person everyone depends on.

Customer Satisfaction

Customer satisfaction measures how callers rate the interaction, usually with a one-question survey on a 1 to 5 scale sent by text or played after the call. CSAT is the share of responses rated 4 or 5. The GSA guidance cited above sets a goal of 4 or 5 ratings on 90% of interactions.

Keep the survey to one question asked right after the call, and add an optional comment. Low scores with comments are worth more than the score itself, because they tell you which part of the call failed.

Agent Occupancy and Call Volume by Hour

Occupancy is the share of logged-in time an agent spends actually handling calls, including wrap-up. Very high occupancy for long stretches leaves no breathing room between calls and tends to show up later as burnout and rising AHT. Very low occupancy means you are staffed for a peak that is not happening.

Both depend on knowing when calls arrive. Pull call volume in 30-minute intervals for at least four weeks, then look for the daily and weekly pattern. Most small teams find two or three predictable peaks they can cover by shifting lunch breaks or start times rather than hiring.

Starting Targets for Call Center Metrics

Benchmarks are a starting point, not a verdict. Use the targets below for your first quarter of measurement, then set your own based on your trend and your customers' expectations.

Metric	Starting target	Source or basis
First call resolution	71% average; 90% as a stretch goal	SQM Group 2021 benchmark; GSA CoE goal
Service level	80% of calls answered within 20 seconds	Industry 80/20 standard (SQM, GSA)
Call abandonment rate	Under 3%	GSA CoE goal
Customer satisfaction	4 or 5 rating on 90% of responses	GSA CoE goal
Average handle time	Your own baseline by call type	No reliable universal benchmark
Transfer and hold time	Trending down month over month	Internal baseline

If you are not hitting these yet, you are in good company: SQM's finding that few centers consistently meet 80/20 is a reminder to judge progress by direction, not by one bad week.

Most of these numbers come straight out of a phone system that has queues and reporting built in. Dial Raven's Pro plan includes call recording and analytics, and our cloud call center tools add queues, skills-based routing and a live supervisor dashboard showing queue volume, wait times and agent status. You can compare plans on our pricing page to see what each tier reports.

How Call Center Metrics Affect Each Other

Metrics are linked, and pushing one in isolation often damages another. These are the trade-offs to watch:

• Cutting AHT by rushing calls lowers first call resolution and raises repeat calls, so total workload goes up
• A strong ASA with high abandonment means callers are giving up, not being served quickly
• Very high occupancy improves cost per call in the short term but tends to raise AHT and turnover later
• Lowering transfer rate by having agents attempt everything can lengthen hold times
• A rising FCR often comes with slightly longer calls, which is usually a good trade

A simple rule for small teams: manage to outcome metrics, first call resolution and customer satisfaction, and use speed and effort metrics to diagnose why outcomes are changing.

How to Track Call Center Metrics Without a Data Team

You do not need a business intelligence tool. A cloud phone system with queues already logs the raw data: when each call arrived, how long it waited, who answered, how long it lasted and whether it was transferred, held or abandoned. The job is to turn that into a routine.

• Put inbound sales and support lines into named queues so each line reports separately
• Turn on wrap-up or disposition codes so agents tag each call as resolved, follow-up or transferred
• Exclude short abandons under about 5 seconds and write the rule down
• Pull a weekly report of abandonment, ASA, service level, FCR and AHT by queue
• Listen to three to five recorded calls a week, picked from the longest and the transferred ones
• Change one thing at a time, such as a schedule, a menu option or a script, and watch the next two weeks

Example: A Five-Person Support Line

Here is a worked example with illustrative numbers. A five-agent support team takes 400 queued calls in a week. Twenty callers hang up after more than 5 seconds in queue, an abandonment rate of 5%. Of the 380 answered calls, 300 are picked up within 20 seconds, a service level of about 79%. Repeat-call tracking shows 95 of the answered callers called back within seven days, so first call resolution is 75%.

Volume by interval shows most of the abandoned calls fall between 11:30 and 1:30, when two agents are at lunch together. Staggering lunches is a free fix to try first. The 95 repeat callers point to a second fix: the three most common callback reasons go into a shared answer sheet, and agents check it before ending a call.

Frequently Asked Questions

Good call center metrics turn a busy phone line into something you can manage: you see where callers wait, why they call back and when you need more hands. Start with abandonment, speed of answer and first call resolution, review them every week and change one thing at a time. If your current phone system cannot show you these numbers, get started with Dial Raven and our team will set up your queues and reporting so you can see them from day one.