Count Logged agents

[cavagnaro]

Hi guys,
Does anyone has an idea on how i can get the average amount of agents per hour logged in?
Also would like to know on real time how many agents i have logged in avoiding to count on CCPulse the number.
Any ideas are welcome
Thanks 

[pdavid]

In the groups view, the way I did this was have the number of folks in each phone state (waiting,internal, inbound, outbound, consult, Not Ready codes) and added a custom formula that adds these up.

This way I can see a total of agents logged in, and if not in 'waiting' mode, I can view how many are on each phone state mode.

Not sure about the average per hour calculation, perhaps someone can chime in with a solution.

[tony]

Search this Forum for "login table".... 

Tony

[cavagnaro]

Yup sorry i did it but was looking more if any new ideas raised up 
I'm evaluatind and testing the Statistical Type and the Login Table info.

[Kevin S]

One method I saw elsewhere is to take the total number of seconds agents are logged in and divide it by the total number of seconds in that time period.

For example, if you have a 1 hour period (3600 seconds) and two agents, one logged in for the entire hour and one logged in for exactly 30 minutes, you would have total log in of 90 minutes login time or 5400 seconds, and 5400/3600 = 1.5 agents.

[mark]

Thats a good way Kevin, I could use that for my GVP ports!

[Kevin S]

Mark -
Per my other post, you may want to use Total Talk Time (or Total Inbound call time, or whatever) instead of Login time for GVP ports. I'm going on the assumption that, except in rare circumstances, the GVP ports will always be logged in, so you would again consistently get the same number (that is, the number of your GVP ports). 

[mark]

Yeah, I created a new placegroup template earlier with some talk time stats. Average may be good enough, its more for licensing than anything!

[pdavid]

[quote author=Kevin S link=topic=2361.msg8782#msg8782 date=1184331150]
One method I saw elsewhere is to take the total number of seconds agents are logged in and divide it by the total number of seconds in that time period.

For example, if you have a 1 hour period (3600 seconds) and two agents, one logged in for the entire hour and one logged in for exactly 30 minutes, you would have total log in of 90 minutes login time or 5400 seconds, and 5400/3600 = 1.5 agents.
[/quote]

I'm trying this out but for I'm missing something, I have a group total that has 68 hours and 27 minutes, so when I convert them to seconds, I get 246,420 and then divide by 3600, I get 68.45 - that group only has 11 agents in it. I know I'm missing something, can someone please help me out with this.

[Kevin S]

Your denominator needs to be the number of seconds in the time period you are calculating for.

Since you said you have 11 agents but 68+ hours, it sounds like you are trying to calculate for the day, not the hour.

In that case, take the 68h 27m (246,420 seconds) and divide by 86400 (24 hours in a day * 3600 seconds per hour = 86400 seconds) and you get 2.85 agents.

[Kevin S]

After thinking over the weekend, unless your call center is 24x7 (with 11 agents, I would be surprised if it was), but you might want to adjust your calculation on your reports for the hours your call center is staffed.

For example, if your call center was only open 8 AM - 6 PM (10 hours), you might want to change the example calculation from the previous post to
(Total Login Seconds)/(hours open * 3600 seconds per hour) ==>
246420 / (10 * 3600) ==> 6.845 agents

I suggest this because, (using my example) if you are not open for 14 of the 24 hours (and therefore your agents are not logged in) and you try to calculate based on the total number of seconds in a 24-hour day, those 14 hours are going to pull your average agent count WAY down.

[cavagnaro]

Hi guys,
I'm using AverageAgentsLoggedIn stat type and seems to work:

[AverageAgentsLoggedIn]
Category=AverageOfCurrentNumber
Description=Average Number of Agents on a time Window
MainMask=LoggedIn, Active
Objects=GroupAgents
Subject=DNAction

Can someone try to implement this and compare with their results?

a problem i have found with this is that on ETL it doesn't create hour/week/month/ ..... reports, only disgregation one. No errors on DBServer or ETL yet found. Is it supposed to work like that?

Thanks

[pdavid]

[quote author=Kevin S link=topic=2361.msg8835#msg8835 date=1185200591]
After thinking over the weekend, unless your call center is 24x7 (with 11 agents, I would be surprised if it was), but you might want to adjust your calculation on your reports for the hours your call center is staffed.

For example, if your call center was only open 8 AM - 6 PM (10 hours), you might want to change the example calculation from the previous post to
(Total Login Seconds)/(hours open * 3600 seconds per hour) ==>
246420 / (10 * 3600) ==> 6.845 agents

I suggest this because, (using my example) if you are not open for 14 of the 24 hours (and therefore your agents are not logged in) and you try to calculate based on the total number of seconds in a 24-hour day, those 14 hours are going to pull your average agent count WAY down.

[/quote]

Hey Kevin,

Thanks alot, this looks to have helped out tremendously, I actually used 12 hours (9am - 9pm), but with in my call center I have different departments that vary on hours of ops, so I'll just need to build templates including this specific per department, which is not a big deal at all.


I currently have 137 hours total, and am getting 11.42 average advisors logged in..Good Stuff..Thanks again.