Average Speed of Answer
Definition
Average Speed of Answer
Average speed of answer (ASA) is the mean time a caller waits in queue before an agent picks up, counted across answered contacts only. Abandoned calls never enter the sum, so the number describes the people who stayed, not everyone who tried.
Outsource Accelerator covers this metric twice already. The general entry sits at average speed of answer (ASA), and the channel-specific one at inbound calls average speed of answer.
Read both for the basics. This page is about what buyers get wrong when a vendor puts ASA on the front page of a reporting pack, because averages hide their own tails.
Key takeaways
- ASA counts answered contacts only, so abandoned calls vanish from the number entirely.
- A mean gets dragged by a few very long waits while most callers were fine.
- Service level, meaning X percent answered within Y seconds, is the better contractual target.
- Where the clock starts — IVR exit or first ring — decides whether two vendors are comparable.
How it works
ASA divides total queue wait time by the number of answered contacts in a period. Nothing else enters the formula. Hold time mid-call, wrap time, transfer time and abandoned waits all sit outside it, which is why ASA moves on its own.
The clock start trips people up. In most platforms the timer begins when the contact enters the agent queue, which happens after the caller finishes with the interactive voice response menu. It rarely begins at first ring.
That boundary is the single biggest source of vendor-to-vendor incomparability. A provider with a long, branching menu can report a flattering ASA while callers press buttons for a minute before the measured wait even starts.
So ask two questions before you compare anything. Where does the timer start, and what happens to abandoned contacts?
Two vendors both quoting 30 seconds can be running completely different experiences for the person holding the phone.
| Element | Inside ASA? | Why it matters |
|---|---|---|
| IVR menu navigation | Usually no | Long menus hide real caller wait |
| Queue wait before answer | Yes | This is the whole measurement |
| Abandoned contacts | No | Removes the worst waits from the mean |
| Hold time during the call | No | Belongs to handle time, not queue |
| Callback wait | Depends on platform | Ask, because treatment varies widely |
Abandonment is the first thing to read alongside it. When waits stretch, more callers hang up, and every hang-up quietly removes a long wait from the calculation.
So a queue that is getting worse can post an improving ASA. That is the trap, and it catches experienced buyers.
Occupancy is the other half of the picture. High occupancy means agents keep almost no idle time, and idle time is exactly what absorbs a spike in arrivals.
Load the roster too tightly and the queue lengthens and spikes. Loosen it and the queue calms down while cost per contact climbs. ASA sits in the middle of that tug of war.
Note what callers actually see, too. Most platforms announce expected wait time, a forward-looking estimate for the person waiting right now, not the historic average in your report.
The reporting window matters just as much. ASA computed across a full month smooths over the two hours a week when the queue genuinely broke.
Pull it by interval instead. Half-hourly is standard — and the weak points show up immediately. Solid workforce management reporting does this without being asked.
None of this makes ASA useless. It is still the fastest read you have on whether a queue is drifting week to week.
That is the honest summary of the metric. Use ASA to diagnose, use a threshold-based target to contract, and never accept one without the other.
Examples
Three settings show the metric behaving differently. Each one explains why the headline figure needs a distribution behind it before you sign a contract, escalate an issue, or hand the number to a board.
Outbound compliance runs a different clock. The FTC’s Telemarketing Sales Rule sits at 16 CFR part 310.
It treats an outbound call as abandoned when a person answers and the telemarketer fails to connect them to a sales representative within two seconds of that completed greeting.
The safe harbour at 310.4(b)(4) protects a seller or telemarketer who keeps abandonment at no more than 3 percent of all calls answered by a person.
That gets measured over a single campaign running under 30 days, or separately over each successive 30-day period. The rule also requires letting the phone ring at least 15 seconds or four rings first.
Note the direction of travel. That two-second standard is an outbound obligation, not an inbound ASA target, and briefs confuse the two constantly.
Public-sector service delivery. The US federal customer experience programme tracks quality across designated High Impact Service Providers.
Executive Order 14058, signed in 2021, established ongoing accountability for federal service delivery and directed 17 agencies to take 36 specific actions on customer experience.
Annual guidance sits in OMB Circular A-11, Part 6, Section 280. Queue speed is one input among many there — the framing is the whole service journey, not a single stopwatch reading.
Staffing economics sit behind the number. The US Bureau of Labor Statistics puts the median hourly wage for customer service representatives at $20.59 in May 2024.
Employment in the occupation is projected to decline 5 percent from 2024 to 2034, with about 341,700 openings a year on average across the decade.
Read those together. Pay is real money and hiring is not getting easier, so every second you shave off ASA has to be bought from a tighter labour pool.
So the queue number belongs in a staffing conversation, not only a quality one. Buy a faster answer and you are buying hours — this metric is a budget line wearing a service badge.
Related terms
ASA never travels alone. Read it beside the metrics below and the picture stops being a single number and starts being a queue you can actually staff, manage and hold a vendor accountable for.
- Average Speed of Answer (ASA): the general OA entry on the same metric.
- Service Level: the percentage of contacts answered inside a fixed time threshold.
- Abandon Rate Percentage: the share of callers who hang up before an agent answers.
- Occupancy Rate: the portion of logged-in time agents spend handling contacts.
- Queue Time: the individual wait a single contact experiences before connection.
FAQ
Why is service level a better contract target than ASA?
Service level counts the share of contacts answered inside a fixed threshold, so it constrains the tail directly. ASA can look perfectly healthy while a minority of callers waits several minutes.
Does ASA include abandoned calls?
No, abandoned contacts are excluded from the calculation. That means the longest waits often leave the number entirely, so always read ASA next to abandon rate.
Where does the ASA clock usually start?
In most platforms the timer starts at queue entry, after the caller finishes with the IVR menu. Get that confirmed in writing, because the alternative reading shifts the reported figure a long way.
How does occupancy affect ASA?
Push occupancy high and queues lengthen, because there is no idle capacity left to absorb arrival spikes. Push it low and ASA improves while cost per contact rises.
Can a vendor improve ASA without improving service?
Yes, since anything that pushes more callers to abandon or lengthens the menu before queue entry can flatter the average.
Compare providers on how they report queue metrics, not just the headline figure, in the Outsource Accelerator directory.
Independent
