KPIs and metrics

Answer Rate and Abandonment Rate: Definitions and Formulas

How to calculate answer rate and abandonment rate, which definitions change the result, how to read them with wait times, and what to do about abandoned calls.

Call editorial team · Published · 5 min read

Two numbers appear on almost every call-centre report: how many calls were answered and how many callers gave up. They sound simple, and they are the first figures a manager is asked about. But they are also the figures most often calculated in different ways by different people, which leads to arguments that have nothing to do with caller experience.

A supervisor reviewing call statistics on a monitor in a contact centre
Illustrative image

This guide defines both measures, shows the formulas, lists the choices that change the result, and explains how to read them. It builds on Contact Center KPIs Explained and sits alongside Average Speed of Answer and Service Level 80/20.

The two definitions

Start with the counts. Offered calls are calls that reached your system and asked for service. Answered calls are those picked up by an agent or by an automated agent that dealt with them. Abandoned calls are callers who hung up after entering the queue, before anyone answered.

The two measures

Answer rateAbandonment rate
Question it answersHow much of the demand did we serve?How many people gave up on us?
FormulaAnswered calls divided by offered callsAbandoned calls divided by offered calls
Good directionHigherLower
Main driverCapacity and routingWaiting time and patience
Blind spotDoes not show how long people waitedDoes not show people who never entered the queue

Multiply either fraction by one hundred to express it as a percentage, which is how most reports show it. If offered calls were two hundred and abandoned calls were ten, abandonment is ten divided by two hundred, or five in every hundred. The numbers here are purely for illustration.

Choices that change the result

Two teams with identical behaviour can report different figures because they made different choices. Decide each of these once, write them down and apply them everywhere.

ChoiceWhy it matters
What counts as offeredInclude calls that reached the menu only, or only those that entered a queue?
Short abandonsCalls that hang up within a few seconds are often excluded, because the caller never really waited.
VoicemailDoes leaving a message count as answered, abandoned or its own category?
CallbacksIf a caller accepts a callback, is the original call answered, abandoned or deferred?
Automated handlingDoes a call resolved by an automated agent count as answered?
Time windowPer half hour, day, week or month? Averages hide peaks.

The treatment of short abandons deserves a note. A caller who hangs up in a few seconds may have dialled the wrong number or changed their mind. Counting those makes abandonment look worse than the queue deserves, so many centres list them separately and leave them out of the rate. Whatever you choose, apply it consistently and show how many were set aside.

Channel mix is another quiet influence. If part of your demand is now answered by an automated agent, or moved to chat and email, the calls that remain for people may be longer and harder, even as the rates look better. Note when you change the setup, and mark the date on your charts, so that a later reader can tell whether a movement was caused by caller behaviour or by a change in how calls are routed.

Reading the two together

Neither figure tells the story alone. A high answer rate with long waits means callers are served, but late. A high abandonment rate may point to a short peak, not a general shortage. Looking at the pattern across the day is more revealing than one number.

A way to investigate a poor result

  1. Break it downLook at each half hour of the day, not the average.
  2. Find the bad intervalsNote when abandons cluster.
  3. Check the waitCompare with the longest wait and the average speed of answer.
  4. Look at the causeStaff on shift, call volume, a long call, a menu problem.
  5. Choose a remedyAdjust staffing, shorten the menu, add a callback or a message.
  6. Measure againCheck the same intervals next week.

Remember the relationship with patience. Callers abandon sooner when they have no idea how long they will wait, when the hold message repeats, or when a callback is not offered. Fixing these can lower abandonment without adding a single agent. IVR Queue Messages and IVR Callback and Voicemail give practical options.

What to do about abandoned calls

A measure is a prompt for action. Every abandoned call is a person who wanted something and did not get it. Some will call again. Others will not. Where you hold the number, a callback is often the most effective answer.

Targets and fairness

Setting a target is a judgement about your business and your callers. There is no universal right answer, because acceptable abandonment depends on the type of call, the cost of a missed one and the alternatives open to the caller. A clinic and a retailer will not choose the same figure. Treat published benchmarks with caution unless they cite a clear source and a clear definition.

Use targets to start conversations, not to punish. Teams that are blamed for a high abandonment rate often respond by changing how it is counted, rather than how calls are handled. See KPI Targets: How to Set Them for a gentler approach.

Checklist for reliable rates

  • Write the definition of offered, answered and abandoned
  • Decide how short abandons, voicemail and callbacks are treated
  • Report by half hour as well as by day
  • Show the number of calls behind each rate
  • Pair the rates with longest wait and speed of answer
  • Review abandoned calls and call back where possible
  • Revisit definitions whenever you change the phone setup

Finding the figures in Call

In Call, the Reports page has an Intervals tab that shows, half hour by half hour for a chosen day, calls offered, answered and abandoned, with short abandons listed but left out of the abandon rate, and average speed of answer and average handle time. It also projects service level by queue. Each queue can have its own target for how many calls to answer within a set number of seconds, and alert limits for the longest wait, the number waiting and the abandon rate.

Callers who hung up while waiting can be added to the callback list, so staff can ring them back, and a card shows how many callbacks were completed on time over seven days. Wallboards can show waiting calls, longest wait, service level and abandon rate on the floor, and reports can be built and exported as a CSV file.

Frequently asked questions

What is the difference between answer rate and abandonment rate?

Answer rate is answered calls divided by offered calls. Abandonment rate is calls that hung up while waiting divided by offered calls. They are related, but other outcomes such as voicemail can sit outside both.

Should very short abandons be counted?

Many centres leave out calls that hang up within a few seconds because the caller never really waited. Choose a rule, apply it consistently and show how many were excluded.

Is there a standard good abandonment rate?

No single figure suits every business. It depends on the type of call, the cost of losing it and the options callers have. Be sceptical of benchmarks without a stated source and definition.

Does a callback count as an answered call?

It depends on your definition. Decide whether a call that is converted to a callback counts as answered, abandoned or its own category, and keep it the same over time.

Why do we need to look at half hours?

Averages hide peaks. A day with a good overall rate may include an hour in which many callers gave up. Intervals show where to act.

How can we lower abandonment without more staff?

Tell callers the likely wait, offer a callback, shorten the greeting and menu, and review the points where callers hang up.

Track abandoned calls and callbacks in Call

In the demo console you can try incoming calls in a simulation.

Related guides