Measuring
How to calculate call abandonment rate (and why two tools give you two numbers)
By the V4N team · Published · 3 min read
The short answer
Call abandonment rate is the share of calls that hang up before an agent answers: abandoned calls divided by offered calls, times 100. The formula is simple; the number is not, because tools differ on whether to count very short hang-ups, calls that reached voicemail, and hang-ups inside the caller menu. Fix the definition first, then look at the rate by hour and by how long callers waited.
Abandonment rate
abandoned calls ÷ offered calls × 100
Offered calls are every call that reached your queue, answered or not.
If your phone system says 6% and your spreadsheet says 9%, nobody is necessarily wrong. They are usually counting different things. Before you set a target or compare yourself with a benchmark, settle these four questions.
Four choices that change the number
- Short abandons. A caller who hangs up after three seconds usually dialled wrongly or changed their mind. Many centres leave hang-ups under a threshold (often 5 to 10 seconds) out of the count. Pick a threshold and write it down.
- Caller menu hang-ups. A caller who leaves during the menu, before ever reaching a queue, may or may not count as offered. Counting them measures your whole front door; leaving them out measures your queues only.
- Voicemail and callback. If a caller leaves a message or accepts a callback, did they abandon? Most teams count them separately so the queue figure stays honest.
- Outside business hours. Calls that hit the after-hours message were never offered to an agent. Keep them out of the rate and report them as their own number.
A worked example
One morning, 1,000 calls reach your queue and 120 of them hang up before an agent answers. Of those 120, 30 hung up in under five seconds.
| Method | Calculation | Rate |
|---|---|---|
| Count every hang-up | 120 ÷ 1,000 | 12.0% |
| Ignore hang-ups under 5 seconds | (120 − 30) ÷ (1,000 − 30) | 9.3% |
Same calls, a gap of nearly three points. If one report uses each method, you will argue about the staffing when you should be agreeing on the definition.
What counts as a good rate?
Published guides commonly cite 5% to 8% as a typical call-centre range and 10% or more as high, but sources disagree on exactly what they count, which is the point of the section above. Treat benchmarks as a rough guide. Your own trend, measured one way every week, tells you more than comparing yourself with a figure from another country and another definition.
The daily average hides the problem
A day at 7% can contain an hour at 25%. Members and customers call at lunch, after payday and before the branch opens, and a flat daily figure smooths that away. Look at the rate for every half hour, laid over the number of agents who were signed in. The gap between the two lines is your staffing problem.
Look at how long they waited
The time a caller waited before hanging up tells you what to fix:
- Most hang up within 10 to 20 seconds: the menu is too long, the first message is too slow, or the line rings too long before the queue starts.
- Hang-ups cluster after 60 to 90 seconds: you are short of agents at that hour, or callers are being sent to the wrong team first.
- Hang-ups are spread evenly: the problem is likely general capacity, and a callback offer will help more than a faster menu.
What to do about an abandoned call
An abandoned call is a customer who wanted something and did not get it. The rate matters less than what happens next. Put every abandoned call, with its number, on a list that someone owns, and call back. We cover that process in how to follow up missed calls.
Then change the cause: staff to the hourly curve, shorten the menu, and offer a callback after a set wait. Check the effect against the same definition you started with.