- This topic is empty.
9th September 1999 at 15:37 #32458Steven ChiaGuest
Given the average number of calls per day, is it possible to estimate the number of calls during peak hour?
For example, if we go by the equal distribution approach, then, over a period of 8 working hours per day, the calls per hour would be = (ave no of calls per day)/8
But in reality, calls are not evenly distrubted. What would be the best statistical model to estimate the peak hour calls. (Poison/Binomial distribution?) How to calculate?
Thanks in advance.9th September 1999 at 15:38 #32459Robert ShiersonGuest
Personally, I’m not sure about the Binomial dist etc but we use an hourly call monitoring package from Sension called MIS, and this gives a call distrib per hour. Depending on what day it is, and what market you’re in (we’re in retail and we get called by travel agents dealing with the public)things can vary greatly esp around the beginning, lunchtime and end of the day. If you need a hourly % Distribution let me know.12th September 1999 at 15:39 #32460BetoanaGuest
Yes, it is.
A sample mode of to calculate the number of calls in a peak hour is aplicating concentration factors. In Fixed telephonic Networks this value is typically 17%. In Mobile Telephonic Networks is a litle diferent, the value is between 8% to 12%. A tipically value is 10%. For example: were carried 100.000 fixed calls during the day, then for classical values of measururement statistics in a peak hour were carried 17.000 calls. Obs. They only statistic values, for the general cases we can to adopt.14th September 1999 at 15:40 #32461Steven ChiaGuest
Thanks, this is certainly useful tips.
But may I know how is this 17% value derived?