How Do We Deal With Outliers Affecting AHT?

Average Handing Time is Heavily Distorted By the Outliers. How Do We Deal With These Outliers

I have read the article https://www.callcentrehelper.com/how-is-average-handling-time-distributed-it-is-not-how-you-think-97688.htm

I do not know how to deal with these outliers which distort the average handing time. By the way, what criteria should you use to identify these outliers? +/- 6 sigmas?

Question asked by john

Use Median for Average Rather Than Mean

The first question is – How are you using this information – and where is this a problem?

If you are using it for Erlang Calculations then this should not be a problem as it is already built into the calculations.

If you are using it for internal reports to show how long the average call lasted, it may be better to show the median, rather than the mean. i.e. 50% of calls had a longer call duration than the median.

It all depends on how you are using it.

With thanks to Jonty

Author: Jonty Pearce