Can anyone tell what would be ideal benchmark for intraday forecast.

Regards,
Vani

I'm guessing you mean accuracy.

<= 5%

If it's anything else let me kow.

HTH

DaveA

Thanks Dave....
You got my query right.

A few more doubts.
In our organisation we are confused whether we should go for 15 minute / 30 minute interval accuracy measurement ?? For most of our product lines the AHT is around 300 seconds.

In a given interval if the call volume is say three calls and we don't get any calls. Then acuracy would be 0%. In such scenarios do you think Target of <=5% would be appropriate??

Regards,
Vani

Vani,

In our organisation we are confused whether we should go for 15 minute / 30 minute interval accuracy measurement

Forecast using 15 minute data for 30 min intervals.

So the rota and forecast that goes out is x calls in 30min even though
the data used to derive the forecast was from 2x15 minute intervals.

In a given interval if the call volume is say three calls and we don't get any calls. Then acuracy would be 0%.

Correct. The forcast is wrong. However you CANNOT be measured on a single interval. The accuracy should be measured over the t component of the forecast.
Expecting 5% accuracy over every interval is err...[1] unrealistic.

Remember 3 calls with a 300sec AHT does NOT mean one agent for that 1/4 hr.
My Qcal says 2-3

HTH

DaveA



[1] You do the math 5 % in EVERY 1/4 hr interval over an 8 Hr shift
gives? Anyone? [2]

[2] I'm not doing all your work for you, but believe me this is a VERY
persuasive arguement.

I would add a caveat to Dave's <=5% and say that if you volumes are very low - e.g. 3 calls ibn 15 minutes = 12 cph = 120 calls in a 10 hour day - then you can expect a far greater variation.

5% of 3000 calls is 150 calls
5% of 300 calls is 15 calls

Clearly 5% in the 2nd example would require extreme accuracy!