Call Abandonment Rate: The Help Desk Number Your Customers Feel First

Call abandonment rate is the percentage of callers who hang up before an agent answers, and a healthy help desk keeps it near 2% with 5% as the ceiling it can live with. It’s the KPI to put on the wall first, because it counts the customers you lost before you ever had a chance to help them. Every other phone number you track describes people who got through.

I’d argue it’s also the most honest number on the report. Average handle time can be flattered and satisfaction surveys only reach callers who stayed on the line. An abandoned call is a vote cast by someone who never got to speak to you.

Why should you track call abandonment rate?

Because the caller who hangs up doesn’t disappear. They call back, which inflates your volume and makes the next interval harder to staff. They send an email that says “I’ve been trying to reach you all morning”, which starts the conversation on the wrong foot. Or they give up on you entirely, and you never find out. The best help desk measures describe an outcome the customer would recognise, and few outcomes are easier to recognise than a phone that rings and rings.

It also exposes a staffing problem faster than any other number. Wait times can look acceptable on an average while one bad half hour sends dozens of callers away. Abandonment is where that half hour shows up first.

Put a price on it with a small example. Say your desk handles 1,000 eligible calls a day and abandons 5%. That’s 1,000 x 0.05 = 50 callers a day who walked away. If, say, 40% of them redial, you take 50 x 0.40 = 20 extra calls tomorrow that you didn’t plan for, and the other 30 are customers you may never hear from again. Those percentages are made up for illustration, so swap in what your own desk sees. The shape of the maths is the point: a small percentage of a big number is a lot of people.

How do you calculate call abandonment rate?

The basic formula is abandoned calls divided by calls offered, multiplied by 100. Call Centre Helper gives the simplest version: offered minus handled, divided by offered, and says the data comes from your ACD abandon report. Count abandons from the network side, including callers who hang up inside the menu, not only those who gave up in the queue.

Then deal with short abandons. Some callers hang up within a few seconds because they dialled the wrong number or redialled to be safe, and counting them punishes you for something no staffing level could fix. The usual convention is to exclude calls abandoned in the first five seconds, and most ACD systems let you set the threshold.

Here’s an illustrative day, and every figure in it is invented so you can follow the arithmetic. The phones offered 1,020 calls and 70 were abandoned, 20 of them inside five seconds.

1. Without the exclusion: 70 / 1,020 = 0.069, which is 6.9%.
2. Take the 20 short abandons out of the top: 70 – 20 = 50 abandons that count.
3. Take them out of the bottom too, because those callers were never really “offered” to an agent: 1,020 – 20 = 1,000 calls that count.
4. Divide: 50 / 1,000 = 0.05, which is 5.0%.

TechTarget’s version of the modified formula works the same way. Pick one definition, write it down and never change it quietly. Moving the threshold from 5 seconds to 20 can improve the report without improving a single call. ICMI’s own KPI material makes a related point: abandonment is influenced by factors the centre can’t directly control, so read it alongside service level and never alone.

How do you track it in Excel?

Make a workbook called Abandon-Tracker.xlsx with one sheet named Intervals. Row 1 holds the headings and each row below it is one interval of the day. You’ll fill in four columns from your phone system’s abandon report and let Excel work out the rest:

• Column A, Interval: the label, such as “Opening” or “Mid-morning”.
• Column B, Offered: every call that reached the system in that interval.
• Column C, Abandoned: every call that hung up before an agent answered.
• Column D, Short: the abandons that happened inside your five-second window.
• Column E, Rate: the formula, which you format as a percentage.

In cell E2 type =(C2-D2)/(B2-D2). Read it from the inside out. C2-D2 is the abandons that count, because it takes the short ones off the total. B2-D2 is the calls that count, because it takes the same short ones off the offered number. The slash divides one by the other, and the percentage format multiplies by 100 for display. For the whole day, put the totals in row 7 with =SUM(B2:B6) and so on down each column, then copy the formula across. With B7 = 1,020, C7 = 70 and D7 = 20, the cell returns (70 – 20) / (1,020 – 20) = 50 / 1,000 = 0.05, which shows as 5.0%.

Don’t stop at the daily total. Break the day into parts, because the average hides the problem. Here are the same 1,000 eligible calls split across the five rows of the sheet. Each row’s rate is its own abandons that count divided by its own calls that count, so Mid-morning is 19 / 250 = 0.076, or 7.6%:

Part of day (column A)	Eligible calls (B minus D)	Abandons that count (C minus D)	Rate (column E)
Opening	150	3	2.0%
Mid-morning	250	19	7.6%
Lunch	200	6	3.0%
Afternoon	250	17	6.8%
Close	150	5	3.3%

Check the totals so you trust the table: 150 + 250 + 200 + 250 + 150 = 1,000 eligible calls, and 3 + 19 + 6 + 17 + 5 = 50 abandons that count. The day looks like a 5.0% problem. It’s really two problems, mid-morning and mid-afternoon, and the fix is about where your people sit in those hours, not about hiring across the board.

One more touch earns its keep. Add a conditional format on column E that turns the cell amber above 3% and red above 5%. After a week of data you’ll see the same rows light up at the same times, and that pattern is your staffing plan telling you where it is wrong.

What target should you set?

Call Centre Helper’s benchmark is about 2% as good and 5% as acceptable, with anything over 5% treated as a problem. Treat those as starting points. A desk where callers are mid-outage will tolerate a longer wait than one taking orders, so set your own target from what your customers will put up with, then hold it for a quarter before you adjust.

When the number is high, the cause is almost always one of three things: too few people logged in during the peak, agents tied up on after-call work, or a menu that makes callers wait before they even reach the queue. Test each one against your sheet. If the red rows line up with lunch breaks, it’s scheduling. If they line up with the minutes after a big call batch ends, it’s after-call work. If the abandons come with a near-zero wait, look at the menu, because callers are leaving before the queue ever sees them. The staffing arithmetic in Erlang C will tell you how many agents a given interval needs, and abandonment tells you whether the answer was right.

One warning about targets. If you put a hard number on abandonment alone, someone will find a way to hit it that doesn’t help callers, such as answering and instantly releasing calls. Pair it with average speed of answer and with the number of tickets that get reopened, so the three numbers keep each other honest.

What should you do this week?

Three small jobs, in this order. First, pull your abandon report for the last five working days and enter it into the five-column sheet above, one row per interval. Second, decide your short-abandon threshold, write it at the top of the sheet, and don’t touch it again this quarter. Third, sort column E from highest to lowest and look at the top two rows.

Those two rows are your answer. Move a break, shift a start time, or put one more person on the phones in that window, then watch the same cell next week. A daily average tells you how you did. The worst hour tells you what to change.