Average Speed of Answer: Formula, Benchmark and a 7-Day Sheet

Average speed of answer (ASA) is the average number of seconds a caller waits from the moment the call joins the queue until a person answers. A queue is the line of calls waiting for a free person. ASA counts answered calls only. Missed calls and calls where the caller hung up first are not in it. Time spent in a phone menu, known as interactive voice response (IVR), is not counted either (Call Centre Helper).

ASA stops the moment someone speaks. How long the conversation then lasts is a separate measure, and average call duration covers it.

Divide the total wait time of all answered calls, in seconds, by the number of answered calls. The result is already in seconds, so there is nothing to multiply.

ASA = total wait of answered calls ÷ number of answered calls

Here is a worked example for a three-person team on one shared number over one week. The figures are invented to show the arithmetic, and the average wait in each band is an assumed figure.

The team received 100 calls and answered 82, so 18 were missed.

• ASA = 984 ÷ 82 = 12 seconds.
• Calls picked up within 20 seconds = 50 + 20 = 70.
• Share of answered calls picked up within 20 seconds = 70 ÷ 82 = 85%.
• Share of all calls received picked up within 20 seconds = 70 ÷ 100 = 70%.
• Answer rate = answered calls ÷ calls received = 82 ÷ 100 = 82%.

A good ASA can hide lost callers because the people who waited longest, or gave up, are missing from the sum. In the example above, ASA is 12 seconds, yet 18 of 100 callers were never answered. ASA cannot show them.

The same blind spot appears across the day. A contact centre can meet a 20-second ASA overall and still have two-minute waits at lunch (Call Centre Helper). An average also hides spread. In a published example, 75 calls had an average wait of 64 seconds and 25 were answered instantly, giving ((75 × 64) + (25 × 0)) ÷ 100 = 48 seconds (Society of Workforce Planning Professionals). Seventy-five of the 100 callers queued, yet the overall ASA reads 48 seconds, and it hides that the long waits sit with the queued callers.

So we read ASA with three companions:

1. Answer rate: answered calls ÷ calls received.
2. Share picked up within your limit, counted against all calls received, not only answered ones (70% in the example, not 85%).
3. The busiest hour, because one bad hour disappears inside a week's average.

For counting callers who hang up on a small line, see call abandonment rate.

A good ASA for a small team is one you set yourself. Our recommendation is a limit of 20 seconds, with 80% of calls answered inside it, because the usual benchmarks describe large contact centres. Call Centre Helper cites a global contact-centre average of 28 seconds but gives no survey, sample or date, and names answering 80% of calls within 20 seconds as the usual target (Call Centre Helper). Neither is an Indian small-team figure.

That 80% target is a service level: the share of calls answered within a time limit. India's telecom regulator sets its customer-care standard the same way. Telecom operators must answer at least 95% of calls with a person within 90 seconds (TRAI Performance Indicators Report, April to June 2026, published 12 October 2026). That rule covers operators' customer care, not a small business, but it shows the regulator measures answer speed as a share within a limit, not as an average.

Our recommendation is to make the share within your limit the target and keep ASA as a read-out. Using both as targets on one queue can make them pull apart, because the two are not mathematically synchronised (Society of Workforce Planning Professionals). Begin with 80% of calls within 20 seconds, then tighten it once you can hit it for two weeks running.

Take the counts from your call history and the timing from ten test calls. The call history gives daily counts of received, answered and missed calls. Stopwatch test calls give the wait time.

On a shared TalkEasy business number, every team member sees the same call history. Our call list shows the caller, the answering agent, call status, timestamps and duration. Sales Call Analytics reports total calls, answer rate, missed calls, average talk time, team performance and peak calling hours, in daily and monthly reports you can download.

The method:

1. Download the week's report and note calls received, answered and missed for each day.
2. Read the peak calling hours across the week and note the busiest one.
3. Time ten test calls after step 2, so you know the busiest hour. Ask someone outside the team to ring the shared number and start a stopwatch when the phone starts ringing. With no phone menu, ringing is when the wait begins. Stop when a person speaks. Spread the calls: two at opening, three in the busiest hour, three at midday, two near closing.
4. Remove the test calls from your received, answered and missed counts, since they appear in the call history too.
5. If your phone system records both the ring start and the pick-up for every call, subtract one from the other and use all answered calls instead of the sample.

Ten calls is a spot check, not a true average. It shows whether pick-up is near your limit, which is the question a small team needs answered.

The sheet has one row per day: calls received, answered, missed, answer rate and how many test calls were picked up within your limit. Fill the counts from the call history and the last column from the stopwatch. The busiest hour of the week goes beneath the table.

Busiest hour of the week: ______

Weekly figures:

• Answer rate = total answered ÷ total received.
• Share within limit = test calls picked up within your limit ÷ 10. Pass = picked up within your limit (20 seconds to start).
• ASA estimate = total seconds across answered test calls ÷ number of answered test calls.

A test call nobody answers within 60 seconds counts as a fail for the share within limit. Leave it out of the ASA estimate, divide by the number of answered test calls, and record it as a missed test call.

When to act (our recommendation):

• Answer rate below 90%: fix missed calls before speed. At that point one call in ten or more goes unanswered, and no ASA will show it.
• Missed calls bunched around the busiest hour of the week: that hour needs more cover.
• Fewer than 8 of 10 test calls within 20 seconds: pick-up is slower than the starting target, so change who rings first.
• After any change, repeat the same seven days and compare.

You lower it with three levers: who rings first, who is free at the busiest hour, and what happens to calls nobody answers.

• Ring order. If the call lands on the person least likely to be free, every call starts with a delay. We let you set call routing rules so incoming calls reach the right team member, and business call forwarding covers how to decide who answers first.
• Peak-hour cover. Use the busiest hours from your sheet and keep breaks and meetings out of them. The cover is staff, not software.
• Callback for missed calls. Callbacks do not lower ASA, because missed calls were never in it. They win back the callers ASA ignores. Our Missed Call Automation is designed to call customers back automatically after missed calls, and the missed call callback workflow shows when a manual task is the better answer.