What’s Really Skewing Your Forecasting

In this instalment of our video series exploring what you need to know about forecasting, and what they won’t tell you, Call Centre Helper’s Jonty Pearce explains AHT and call distribution.

AHT and Call Distribution

Here’s a misconception that a lot of people have about average handling time: so you’ve got an average handling time of five minutes, and a lot of people think that the average handling time of five minutes is distributed sort of like this, something like a normal distribution.

You’ve got a lot of calls around here, and then it sort of tails off accordingly. Unfortunately that is not at all how average handling time is distributed. It’s not even distributed like this. You might have heard it’s a Poisson distribution, so it’s not a lopsided distribution.

What you actually have with average handling time is you have what is called, funnily enough, the Erlang distribution.

So you have your average handling time of five minutes, but actually there aren’t many calls distributed around the average. What you find is a lot of short calls, people who’ve got very simple enquiries and they call in and get those answered very quickly, and what you also have is some very long complex enquires.

What that tends to have an effect on is that you get very skewed results, so for instance although an Erlang calculator will say you need 208 agents, what you’ll actually find is that the average number of agents that you need will vary quite considerably from what the Erlang calculator will say.

The Erlang calculator will tend to give you an average number of people you need, but that could vary by sort of plus or minus 10% quite dramatically. Depending on how many short calls you get in that period, or how many long calls you get in that period, across a long enough time period, it will tend to average out like this and that is called the Erlang distribution.

For more in this series watch these videos next:

• Getting Started With Forecasting
• Is Overhang Distorting Your Staffing Calculations?
• Why Your Forecasters Should Always Talk to Marketing

Author: Jonty Pearce
Reviewed by: Robyn Coppell