Average Abandon Time
Definition
Average Abandon Time
Average abandon time is the mean number of seconds callers waited before hanging up, measured only across the calls that abandoned. It’s a patience metric, not a volume metric, so it tells you how long people tolerated your queue before they finally gave up.
The metric answers a narrow question: among the people who gave up, how long did they last? It says nothing about how many gave up. That’s the job of call abandonment rate.
Report the two side by side and you get a story. One number tells you how badly the queue is bleeding — the other tells you where the bleeding starts.
Key takeaways
- Average abandon time is a duration in seconds, never a percentage.
- It counts abandoned calls only, so a thin sample swings it hard.
- A short average points at the front door: menus, greetings, opening prompts.
- A long average points at staffing, because patient callers still went unanswered.
How it works
Add up the wait time of every call that abandoned, then divide by the number of abandoned calls. Calls that reached an agent never enter the sum. That single exclusion is what makes the number so easy to misread.
The clock starts when the caller enters the queue and stops when they hang up. Time spent in the greeting or the menu may or may not be included, depending on how your platform defines queue entry.
Check that setting before you compare two teams. One queue counting menu time against another that doesn’t will look slower for reasons that have nothing to do with service.
Means are fragile on small samples too. A single caller who forgot they were on hold can drag a quiet queue’s average up by half a minute, and you’d be tuning staffing for one distracted person.
Most platforms also let you set a short-abandon threshold, so calls dropped almost instantly never count. Switching it on lifts the average. Switching it off drops it. Neither choice is wrong, but the two can’t be compared.
Three checks before you trust the number:
- Count the abandoned calls sitting behind the average, since a mean built on a handful of records moves on noise alone.
- Confirm where your platform starts the clock, because greeting and menu seconds can sit inside or outside the recorded wait.
- Check whether a short-abandon filter is quietly removing instant hang-ups from the population you’re averaging.
Once you trust the number, the reading is directional. Short and long point at completely different departments.
| Average abandon time | Most likely cause | First place to look |
|---|---|---|
| Under 20 seconds | Callers bail at the front door | Menu depth, greeting length, opening prompt |
| 20–60 seconds | Expectation gap inside the queue | Queue-position and estimated-wait announcements |
| Over 60 seconds | Patient callers, too few agents | Staffing model, schedule adherence, occupancy |
Treat those bands as a starting hypothesis rather than a verdict. Your own queue may sit somewhere else entirely, and what matters is the direction of the read, not the exact seconds.
A short average is rarely a compliment — it usually means people quit within the first few rings. They never heard a queue-position message and never believed anyone was coming.
A long average flips the diagnosis completely. Those callers waited, trusted you, and still got nothing, which is a coverage problem you solve with headcount and scheduling.
Staffing math treats patience as an input. Erlang C assumes nobody abandons at all — so queues with a long average abandon time drift furthest from what the model predicts.
Callback offers and queue-position announcements both move the number, and they move it in opposite directions. A callback pulls impatient people out of the queue early. An announcement talks borderline callers into holding on longer.
Pair the metric with average speed of answer and you can see both sides of the same wait. One covers the callers you reached. The other covers the ones you lost.
Examples
Numbers only earn their keep when you attach them to a shift, a queue, and a decision. Here’s how average abandon time reads in three ordinary situations a contact center team runs into most months.
A billing queue averaging 12 seconds. Callers heard a long menu, guessed wrong, and quit before the hold music even started. Abandon rate looked awful, yet nobody had waited long enough to blame staffing.
The fix was menu design. Moving billing to the first option and trimming the opening prompt cut the losses without adding a single agent to the roster.
Outbound calls follow a stricter rule. The Federal Trade Commission treats a call as abandoned when nobody connects a live representative within two seconds of their completed greeting.
The full text sits at 16 CFR part 310 on the eCFR, and its safe harbour caps abandonment at no more than 3 percent of all calls answered by a person.
That 3 percent is measured over a single campaign shorter than 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 before you disconnect.
So an outbound dialler team runs two definitions at once. One is a compliance test measured in two seconds, and the other is the operational average their own inbound queue reports.
A support line averaging 90 seconds. Those callers waited a minute and a half and still got nothing. That’s a rostering failure — not a menu problem.
Fixing it costs money. The US Bureau of Labor Statistics put the median hourly wage for customer service representatives at $20.59 in May 2024.
The same source projects employment in those roles to decline 5 percent from 2024 to 2034. So the practical answer often looks like sharper routing and better scheduling rather than simply buying more seats.
Related terms
Average abandon time only makes sense next to the metrics it sits beside. Each of these covers a slice of the same queue behaviour, and reading them together stops you from fixing the wrong end of the problem.
- Abandonment Rate: share of offered calls that end before an agent answers.
- Abandoned Call: single contact that ends before it reaches a live representative.
- Average Speed of Answer: mean wait across answered calls, the mirror image of this metric.
- Service Level: percentage of calls answered inside a stated time threshold.
- Call Queue: holding line that orders waiting callers before routing.
FAQ
Is average abandon time the same as abandon rate?
No. Abandon rate counts how many callers gave up, while average abandon time measures how long they held on first. You need both to know whether the problem is the greeting or the roster.
What counts as a good average abandon time?
There’s no universal target, because the right number depends on your queue design and the promise you made. Read it against your abandon rate and average speed of answer rather than an outside benchmark.
Why did my average abandon time rise after adding callbacks?
Callbacks catch impatient callers before they hang up, so they never enter the abandoned population. The people who still abandon are the patient ones — so the average climbs even while fewer calls are lost.
Does the two-second outbound rule apply to my inbound queue?
No. The two-second standard describes outbound telemarketing under the Telemarketing Sales Rule, and inbound average abandon time is an operational metric with no regulatory threshold attached to it.
Should I report it in seconds or in minutes?
Report it in seconds, since the differences that change your decisions usually sit inside the first minute.
Compare delivery partners who already run queue diagnostics like this one across the Outsource Accelerator hubs.
Independent
