Free tool Customer service metrics calculator
Handle time, resolution, backlog, workload and cost are the five questions every support review comes back to. Pick a group, enter the totals from your helpdesk or chat report, and get the KPI with its definition written next to it — so two teams mean the same thing by the same word.
- Eight core KPIs
- Definitions included
- Runs locally
Pick a group of metrics
Each group has sample totals for one month of a 10-person team. Replace them with your own report figures.
Calls, chats or tickets an agent worked on in the period.
Time spent in the conversation.
Customer on hold or waiting mid-conversation. 0 for chat if not tracked.
Notes, tagging, follow-up tasks after the conversation.
All conversations (or issues) in the period.
Marked resolved by the end of the period.
Solved without the customer having to come back. Optional.
Resolved, then the customer wrote again. Optional.
The current backlog.
Average over the last few weeks.
Average over the same weeks.
In the period.
Two half-time agents = 1.
Sum of handle time across agents.
Ready or handling, from your system's status report.
Contracted hours in the period.
Salaries, software, overhead for the period.
In the same period.
Optional.
Optional.
Average handle time (m:ss)
8:00
(talk + hold + after-contact work) ÷ contacts handled.
Resolution rate
—
Resolved ÷ all conversations.
Days to clear the backlog
—
Open tickets ÷ (resolved per day − new per day).
Conversations per agent per day
—
Conversations ÷ (agents × working days).
Cost per contact
—
Total cost ÷ contacts handled.
Share these inputs: the values go after the # in the link, so they never reach a server.
https://tapcub.com/tools/support-metrics-calculator
Version v1.0Updated Update history
What each group answers
No single KPI describes a support team. Each group answers one question, and each can be gamed if read alone.
How long does a contact take?
AHT, 8:00 in the sample. It drives staffing (see the Erlang C calculator) but pushing it down can push resolution down with it.
Do problems get solved?
Resolution rate 90% and first contact resolution 70% in the sample. The gap is customers who had to come back; reopens show resolutions that did not hold.
Is the queue under control?
A backlog of 600 that shrinks by 60 a day clears in 10 days. If new tickets outpace resolutions, it never clears — hiring or deflection is the only fix.
Are people busy, and what does it cost?
Occupancy 78.6% and utilization 62.5% in the sample, and $5.00 per contact. High occupancy with low utilization means plenty of paid time is spent logged out.
Every one of these depends on definitions — what counts as resolved, which time counts as handling. Write the definition next to the number whenever you report it.
From a monthly report to KPIs
Use totals for one period, from one source, for one team.
-
1
Choose a group
Pick the question you are answering: handle time, resolution, backlog, workload or cost.
-
2
Enter totals, not averages
Use totals for the same period — total hours, total conversations — so the calculator does the division consistently.
-
3
Read the definition
The formula is written under each result. Check it matches the one your team or client uses.
-
4
Compare like with like
Compare a team with itself over time, or two teams with the same definitions and the same channel mix.
Eight KPIs, eight formulas
The resolution formulas follow TapCub's resolution rate definition; the others are the common industry forms, with the variants noted.
- First contact resolution can also be calculated over resolved conversations — both are shown
- Utilization is sometimes defined as logged-in ÷ paid time; that figure is shown as "logged-in share"
- Days to clear assumes the current daily rates continue
- Cost per contact includes whatever costs you put in the total; state which
| KPI | Formula | Sample |
|---|---|---|
| AHT | (talk + hold + after-contact work) ÷ contacts | 160 h × 3,600 ÷ 1,200 = 8:00 |
| Resolution rate | resolved ÷ conversations | 1,800 ÷ 2,000 = 90.0% |
| First contact resolution | resolved on first contact ÷ conversations | 1,400 ÷ 2,000 = 70.0% |
| Days to clear | open ÷ (resolved/day − new/day) | 600 ÷ (260 − 200) = 10 |
| Conversations per agent per day | conversations ÷ (agents × days) | 4,400 ÷ (10 × 22) = 20 |
| Occupancy | handle time ÷ logged-in time | 1,100 ÷ 1,400 = 78.6% |
| Utilization | handle time ÷ paid time | 1,100 ÷ 1,760 = 62.5% |
| Cost per contact | total cost ÷ contacts | $45,000 ÷ 9,000 = $5.00 |
Two teams, same resolution rate
Two support teams in the same month, each with 2,000 conversations. Sample figures.
| Team | Resolved | First contact | Reopened | Resolution rate | FCR | Reopen rate |
|---|---|---|---|---|---|---|
| Team A | 1,800 | 1,400 | 90 | 90.0% | 70.0% | 5.0% |
| Team B | 1,800 | 1,000 | 270 | 90.0% | 50.0% | 15.0% |
Both close 90% of conversations, but Team B needs a second contact far more often and three times as many of its resolutions come back. Resolution rate alone would rank them equal.
Where support KPIs mislead
The arithmetic is simple. The traps are in what goes into it.
Different denominators
FCR over all conversations and FCR over resolved ones can differ by 20 points. Say which one you use.
Driving AHT down alone
Shorter conversations look efficient until customers come back. Watch FCR and reopens when AHT drops.
Confusing occupancy and utilization
One uses logged-in time, the other paid time. Mixing them makes a busy team look idle or the reverse.
Closing instead of solving
Resolution rate set by agents can rise because conversations are closed sooner. Check satisfaction and reopens alongside.
The chat KPIs, without the spreadsheet
For website chat, TapCub's support report calculates the main numbers from every conversation.
Resolution rate
Resolved ÷ all conversations in the range, with comparison to the previous period and resolved conversations per agent.
Learn moreFirst response time
Average and median first response, trend over time and average per agent.
Learn moreChat × analytics
Compare visitors who chatted with those who did not: conversion rate, order value and repeat rate.
Learn moreCustomer service metrics questions
Still have a question?
Chat with the team behind TapCub — we usually reply within a few hours on business days.
What are the most important customer service metrics?
Most teams track speed (first response time, AHT), outcome (resolution rate, first contact resolution, satisfaction), volume (backlog, conversations per agent) and cost (cost per contact). Pick a few from each area rather than many from one.
How do you calculate average handle time?
Add talk or chat time, hold time and after-contact work for all contacts handled, then divide by the number of contacts. 160 hours over 1,200 contacts is 8 minutes.
What is the difference between resolution rate and first contact resolution?
Resolution rate counts every resolved conversation. First contact resolution counts only those solved without the customer coming back. Both are divided by all conversations here; FCR over resolved conversations is shown as well.
What is the difference between occupancy and utilization?
Occupancy is handling time as a share of time logged in and available. Utilization is handling time as a share of paid time. Utilization is always the lower of the two.
How long will it take to clear our backlog?
Divide the open tickets by how many more you resolve than receive each day. If you receive as many as you resolve, or more, the backlog will not clear without more capacity or fewer incoming tickets.
How do I calculate cost per contact?
Divide total support cost for a period — salaries, software, overhead — by contacts handled in the same period. Cost per resolved contact divides by resolved contacts instead, which penalises unresolved work.
Where do first response time and CSAT fit?
They have their own calculators: the first response time calculator for speed and the CSAT calculator for satisfaction.
Update history
New features, fixes and data refreshes for this tool, newest first.
-
· v1.0
LaunchedFirst release.
Missing something or found an error? Tell us — requests that come up often go into the next update. All tool updates
Turn the numbers into a staffing plan
Handle time and volume are the inputs to staffing. Feed them into a staffing model to see how many people each hour needs.