Average Speed of Answer (ASA): Definition, Formula, and Benchmark

Average speed of answer, or ASA, is how long callers waited, on average, between entering the queue and being connected to an agent. It is one of the three standard queue metrics, alongside service level and abandon rate, and the easiest to explain: "callers wait about half a minute." It is also the easiest to misread, because an average of waits says nothing about how many callers waited far longer than the average, or how many gave up.

What Is Average Speed of Answer?

ASA measures the queue, not the agent. The clock starts when a call leaves the IVR and enters a queue waiting for an agent, and stops when an agent answers. Call Centre Helper's guide states the calculation as "ASA = Total Wait Time ÷ Number of Answered Calls," with the example "ASA = 90 seconds ÷ 3 calls = 30 seconds," and specifies that the measure "includes the time a caller spends waiting on hold but excludes time spent navigating an IVR (Interactive Voice Response) system" and should count only "calls that were actually answered, not abandoned or missed calls."

Those exclusions matter. Because abandoned calls are not in the denominator, a queue can post an excellent ASA on a day when a third of callers hung up, since the callers who waited longest are exactly the ones who left. ASA describes the experience of the callers who stayed.

Phone platforms differ on whether ring time at the agent's station counts, and on how transfers are treated. The usual convention is to measure from queue entry to agent answer, including ring time, and to treat a transferred call's second wait as a separate measurement. Whatever the platform does, the definition should be written down next to the number.

How to Calculate ASA

ASA = Sum of queue wait times for answered calls ÷ Number of answered calls

Worked example: a 50-agent inbound team, one interval

Between 10:00 and 10:30, 410 calls were answered. Their queue waits summed to 13,940 seconds. Another 38 calls were abandoned after waiting a combined 4,180 seconds.

• ASA = 13,940 ÷ 410 = 34.0 seconds

Now look at the distribution behind that average. 290 calls were answered within 20 seconds (average wait 8 seconds); 120 calls waited longer, averaging 96.8 seconds. The 34-second ASA is a blend of a large group who barely waited and a smaller group who waited a minute and a half. If the 38 abandoned calls, which averaged 110 seconds before hanging up, had stayed and been answered, ASA would have been (13,940 + 4,180 + additional wait) ÷ 448, well above 40 seconds. The service level for the interval was 290 ÷ 410 = 70.7 percent in 20 seconds; the abandon rate was 38 ÷ 448 = 8.5 percent. Three metrics, one queue, three different impressions.

ASA Benchmarks

Across Call Centre Helper's coverage, 28 seconds is repeatedly described as the global average and the standard figure for this KPI, and the 80/20 service level convention, 80 percent of calls in 20 seconds, is often cited alongside it. Industry commentary generally places a good ASA in the 20- to 30-second range. Call Centre Helper adds the caution that there is so much variety between centres that few rules apply widely, and the ASA guide itself gives no single recommended target.

The benchmark should be read with its conventions. A center that excludes short abandons and IVR time, and measures to the agent's answer, will post a different ASA than one that includes ring time from a different point. And an ASA target set without a service level target invites the wrong behavior: a center can hit 28 seconds by answering easy queues quickly and letting a hard queue wait.

ASA vs Service Level vs Abandon Rate

Metric	Formula	What it hides
ASA	Mean wait of answered calls	The distribution; long waits and abandons
Service level	Percent answered within X seconds	How long the calls outside X waited
Abandon rate	Abandoned ÷ offered	How long answered callers waited

Service level is usually the operating target because it is a threshold and scheduling tools such as Erlang C are built around it. ASA is the reporting metric because it is intuitive. Abandon rate is the check on both.

Why ASA Matters

Wait time is the part of the call the customer notices most and the agent controls least. It drives abandonment, repeat calls, and the mood of the caller when the agent finally answers, which then drives handle time. And ASA is a direct readout of staffing: when scheduled agents are missing from an interval, ASA rises within minutes.

How to Improve ASA

• Staff to the interval forecast. ASA is almost entirely a function of agents available versus calls arriving. Capacity planning and forecast accuracy at the half-hour level are the main levers.
• Protect adherence. Late logins and stretched breaks in a busy interval raise ASA more than any other agent behavior. Schedule adherence is the agent-side metric that moves ASA.
• Route by skill and priority. Sending calls to the first available qualified agent, and prioritizing queues that abandon fastest, lowers ASA without adding staff.
• Reduce handle time where it is waste. Shorter after-call work returns agents to the queue sooner; shorter talk time at the expense of resolution does not help.
• Use callbacks in peaks. A callback converts a long wait into no wait and takes the call out of the ASA calculation entirely, which is fine as long as the callback is honored.

How to Misuse ASA

The most common misuse is treating ASA as a target on its own, which rewards answering the easy queue and starving the hard one. The second is comparing ASA across centers with different exclusion rules. The third is coaching agents to answer faster; agents answer as soon as they are available, and the only thing "faster" can mean is rushing the previous call.

ASA in Contact Centers and Remote Teams

A 50-agent remote billing queue ran a 24-second ASA on paper and a 6 percent abandon rate that did not fit. Looking at the interval data, the 9:00 to 10:00 hour had an ASA of 71 seconds and accounted for most of the abandons; the rest of the day was under 15 seconds, which dragged the daily average down to a number that looked healthy. Attendance records for the 9:00 interval showed three of fourteen scheduled agents logging in after 9:10 most days. The fix was an attendance conversation, not a routing change, and the 9:00 ASA fell to under 30 seconds within two weeks.

How to Track the Staffing Behind It

HiveDesk does not calculate ASA; that figure comes from the phone system. What it provides is the record of how many agents were logged in and available during each interval: scheduling shows the plan, clock-in, clock-out, and break records show the reality, and attendance history shows whether a late-login pattern is one agent or a shift. Matching a rising ASA to a staffing gap in the same interval is how the cause gets found. The plan is $5/user/month with a 14-day free trial and no credit card required. The call center KPI calculation guide walks through the queue metrics together.

Related Terms

• Service level: the threshold version of the same queue measurement
• Abandon rate: the callers ASA leaves out
• Schedule adherence: the agent behavior that moves ASA fastest
• Occupancy rate: the agent-side sign that ASA is about to rise
• Average handle time: how long agents are unavailable between answers

Frequently Asked Questions

What is average speed of answer?

The mean time callers spent waiting in queue before an agent answered, calculated as total wait time of answered calls divided by the number of answered calls. It excludes IVR time and, by convention, abandoned calls.

What is a good ASA for a call center?

The most commonly cited industry figure is 28 seconds, and 20 to 30 seconds is generally described as a good target. Targets vary by sector and by how the metric is defined, and ASA should be read alongside service level and abandon rate.

How is ASA different from service level?

ASA is an average wait. Service level is the percentage of calls answered within a threshold, such as 80 percent in 20 seconds. Service level shows how many calls beat the target; ASA shows the mean wait and hides the distribution.

Does ASA include abandoned calls?

Not by the standard definition, which counts only answered calls. That is why ASA can look good on a day with high abandonment: the longest waits belonged to callers who left.

Does ASA include time in the IVR?

No. The queue clock starts when the caller leaves the IVR and waits for an agent. Some platforms include ring time at the agent's station; the definition should be stated with the number.

Can agents improve ASA?

Only through availability: logging in on time, returning from breaks on schedule, and keeping after-call work short. Agents cannot answer faster than calls are routed to them, and coaching them to "answer faster" usually means rushing the previous call.

Browse more workforce management terms in the glossary.