Hi All,

In absence of no workforce software, I am trying to forecast and analyse call volume. I want to know the following:

1. What my overall new forecast should be for Mondays, Tuesdays etc ( i have all the histrocial data)
2. What or how should I determine what my forecast should be on hourly basis?
3. Would r2 formula be helpful in determining if there is need to change the forecast? If yes; what is the foruma as I cant re-call it?

I have all the historical dat broken by hours and per day.

Appreciate your help in advance. I am trying to do all of this in excel. Its a small company so there is no scheduling or forecasting software.

Thanks :-)

Hi Naomi -

1 - There are different ways of generating forecasting for a month/week. Say suppose you have identified 1000 calls to be your weekly volume

Check the DOW (day of week) allocation and divide these 1000 calls to get daily volume.

2 -
Now check the call arrival pattern at an interval level for different days of a week (you may consider previous 3 weeks data to find the arrival pattern)

Distribute the daily volume as per the call arrival pattern, which will be interval level volume

3 - R2 (coeff of determination) is absolutely not a measure for forecasting accuracy, you may use regression models to predict/forecast and R2 will signify the significance of the variation on the dependent variable because of the independent variable(s).

Forecasting accuracy is rather determined by measures like MAPE, MAD, MSE.

Thanks!
Kaushik

Thts a good answer Kaushik.

Volume accuracy can be determined through different ways

1) No. of instances/days out of day/month/year you were within +/- variance (eg)
2) Simple variance (Actual-Forecast)/Forecast
3) MAPE

Its upto you to determine what works for you best. Finding accuracy/inaccuracy is not oly about knowing it. Its what you do after knowing it makes it worthwhile.