Normalising Calls for Forecasting FTE Requirements

Normalising Calls for Forecasting FTE Requirements

When forecasting inbound call volumes (for the purpose of calculating FTE requirements) should you normalise historical call volumes using the actual abandoned rate?

Let’s assume that your answer rate target is 95%. Last year you received 100,000 calls but your actual answer rate was only 90%.

Should you base your forecast on 100,000 calls or a lower figure, for example 94,737 (90,000/95%)? The argument behind the lower figure is that your poor answer rate artificially inflated call volumes (repetitive callers).

Which figure would you use for future forecasting and why?

Question asked by Dorota

You Should Remove Some of the Abandoned Calls

You should remove some of the abandoned calls from the volumes.

In order to do this you need to look at the propensity to call back from abandoned calls.

Ideally you should do a trace using call detail records. Take the last 3 days of calls and look at 100 calls that abandoned 3 days ago. Look to see if they called back in the last 3 days and use that as a ratio. You then use that ratio to factor in the underlying demand.

If you cannot access these records then a rule of thumb is that for a customer service line 50% of the callers who abandon will call back and the other 50% won’t. In a sales like it is more line 10% that will call back.

With thanks to Jonty

Author: Jonty Pearce