Average Speed of Answer: How to Measure and Improve It
What average speed of answer (ASA) measures, how it differs from service level and longest wait, how to calculate it and how to improve it.
Few numbers feel as personal to callers as the wait before someone answers. Average speed of answer, ASA, puts a figure on it. It is one of the oldest contact centre measures and still appears in most dashboards and contracts.

This guide explains what ASA is, how to calculate it, how it differs from related measures, why averages can mislead and what you can do when it is too high. It sits beside Service Level Explained and Answer Rate and Abandon Rate.
What ASA measures
ASA is the total time that answered calls waited in queue before reaching an agent, divided by the number of answered calls. The clock usually starts when the call enters the queue and stops when an agent answers. Time spent in the IVR menu is normally not part of it, though some reports include it, so check the definition.
Calculating ASA for a period
- Collect queue wait timesFor each answered call, the seconds between entering the queue and being answered.
- Add them upSum the waits for all answered calls.
- Count the answered callsDo not include abandoned calls in the divisor.
- DivideTotal wait divided by the number of answered calls.
- Choose the periodHalf hour, day, week. The shorter the period, the more useful for action.
A hypothetical example: if five answered calls waited 10, 15, 20, 25 and 80 seconds, the total is 150 seconds and the ASA is 30 seconds. Note how the single long wait pulls the average up while four callers waited 25 seconds or less.
ASA and its neighbours
| Measure | What it tells you | What it misses |
|---|---|---|
| ASA | The average wait of answered calls | Spread: a few long waits can hide in a good average |
| Service level | The share of calls answered within a set time | How long the others waited |
| Longest wait | The worst wait right now or in a period | How typical it is |
| Abandon rate | How many callers gave up before an agent answered | Why they gave up or whether they called back |
| Answer rate | How many offered calls were answered at all | How long answering took |
No single measure tells you whether callers are well served. ASA tells you about the typical answered caller; the others fill in the rest. See Call Center KPIs for the full picture.
Why averages can mislead
Common ways ASA misleads
- Survivor biasCallers who abandoned never count in ASA. When waits get long, more people hang up, which can make ASA look better than it feels.
- Skewed by outliersOne very long wait lifts the average while most callers waited briefly.
- Wide intervalsA daily ASA can mix a quiet morning with a painful lunchtime.
- Different definitionsWhether IVR time, ring time or callback time are included varies, so comparisons need care.
What drives ASA
ASA rises when calls arrive faster than agents can take them. That can happen for several reasons:
- Staffing mismatched to arrivals. The right number of agents, but at the wrong hours.
- Long handle times. Agents stay on calls longer than planned, so fewer are free.
- Unplanned absence. Sickness or late starts reduce capacity.
- Surges. A mailing, an outage or a news event brings a burst of calls.
- Skill mismatches. Calls wait for agents with a specific skill while others are free.
- Avoidable repeat calls. Callers ring again because the first call did not solve the problem.
Improvement works best as a loop: measure by half hour, change one thing, then look again after a week or two. Changing five things at once makes it impossible to know what helped.
Improving ASA
Plan people to match calls
Compare scheduled staff with the forecast for each half hour and adjust shifts, breaks and training time. See Workforce Schedules in Call, Forecasting Call Volume and Shift Planning Basics.
Route calls to the right people
Skills-based routing and sensible priorities stop calls waiting for the wrong agent. Keep the number of skills small enough that pools of agents stay large; see What Is ACD? Call Routing Methods Explained.
Reduce demand you can avoid
Analyse reasons for calling and fix the top few: a confusing bill, a missing status update, a broken link. Cutting repeat calls lowers the load for everyone.
Give callers choices to waiting
A callback option lets callers keep their place without holding. A clear message about expected wait helps people decide. See Using Callbacks and Voicemail in Your IVR.
When ASA is too high
- Check which half hours are worst, not just the daily figure
- Compare scheduled staff with forecast for those periods
- Look at handle times for changes
- Check for skill or routing bottlenecks
- List the top reasons for calls and any avoidable ones
- Offer callbacks during peaks
- Review again after a week and record what helped
A worked comparison
Picture two queues with the same ASA of thirty seconds. In the first, nearly every caller waits between twenty and forty seconds. In the second, most callers are answered in ten seconds, but a handful wait several minutes, and a few hang up. The averages match, yet the experience is very different, and the second queue has a problem the average does not reveal.
This is why teams that care about wait time track the spread as well as the mean. A simple way is to look at the longest wait in each half hour and at how many calls were answered within a set time. Another is to plot waits for a day and look for clusters of long ones, which often point to a shift gap, a break scheduled at the wrong time or a skill missing from a rota.
- Compare the median wait with the mean: a big gap suggests a long tail.
- Review the ten longest waits each week and ask what they had in common.
- Check whether long waits cluster at shift changes, lunch breaks or the start of the day.
- Look at callers who abandoned: how long did they wait before leaving?
Callers also judge waiting by what they are told. A caller who hears an honest estimate and an offer of a callback usually tolerates a wait better than one left with silence or music. Make sure the messages you play match reality, and revisit them when waits change.
Setting an ASA target
Some organisations set an ASA target, such as a number of seconds, but it is usually better to set a service level target and watch ASA as a supporting measure. If you do set an ASA target, set it per queue, review it against abandon rate and do not reward teams for keeping ASA low by hanging up on callers or pushing them to voicemail.
Doing it in Call
In Call, the Intervals tab in Reports lists, for each half hour, calls offered, answered and abandoned along with ASA and average handle time. Calls abandoned very quickly are listed as short and left out of the abandon rate. Check the screen's own definitions when you compare with another system. The wallboard shows calls waiting and longest wait as live tiles whose colour changes when a threshold is crossed. On the Thresholds tab of a queue, you can set an alert when the longest wait exceeds a number of seconds. The Workforce page compares required with scheduled staff by half hour, which helps with the staffing side. Callbacks and queue voicemail give callers alternatives to waiting.
Frequently asked questions
What is a good ASA?
It depends on your callers and your channel. An urgent line needs a much shorter wait than a general enquiry line. Set a target by queue, watch it with service level and abandon rate, and judge it against what your callers will accept.
Is ASA the same as service level?
No. ASA is the average wait of answered calls. Service level is the share of calls answered within a set time. Two queues can have the same ASA and very different service levels.
Do abandoned calls count in ASA?
Usually not. ASA is calculated over answered calls. That is why you should always read it alongside abandon rate.
Does ASA include IVR time?
Normally it measures time in queue, from entering the queue to answer. Some reports include other time, so check the definition before comparing.
How can I lower ASA quickly?
Look at the worst half hours and add capacity or move breaks, route calls to the right skills and offer callbacks during peaks. Longer-term, reduce avoidable repeat calls.
Why did ASA improve when customers seem less happy?
If waits got long and more callers gave up, the remaining answered calls may show a lower average. Check the abandon rate and callbacks before celebrating.