Network call reporting - double couting abandons?

[jcarroll01]

I've set up reports for all skills at two sites and report on them separately...however...the abandons seem to be duplicating. Since we basically queue the calls at one site and reserve agents for the calls on the other site should there ever be abandons reporting at the remote sites? How does one setup offer - answered - abandoned in this situation?

 

[Milesprower]

You can get double counts in a networking situation. I'm not sure where it is, but there is some discussion of how this can happen in the documentation.

Basically, if a caller abandons in the period of time where the call is being connected to the far-end reserved agent, it may be counted as abandoned twice (once in the local primary script, and once in the far end network script).

One of the downsides of networking is the increased reporting complexity.

 

[picklemogg]

Yes- you will definitely see abandons double peg from a skillset view in a networked environment. From an application view abondoned calls only peg once. It actually makes sense which is hard to believe since we are talking about Nortel reporting. Example- call is waiting at CS appl at Site A. Through Site A CS appl- CS skill is queued to at sites A, B, and C. After 30 seconds (before an agent in CS skill becomes idle) the customer decides to hang up. Call pegs as abandoned in Site's A CS appl. At skillset view- call pegs as abandoned in CS skill at Site's A, B, and C. So if you would run a skill report from Site A, B, or C you would see this call abandoned.

In our environment we have 6 networked CCMSs. For certain skills we see the call abandon at each site or 6 times. We've had to modify what report tables (appl vs skill) for certain information. Use appl for calls abadoned if possible. If you use skill, divide abandons by number of sites skill is networked to.

 

[jcarroll01]

Can I just zero out the abandons at 2 of the 3 networked sites at the Skill level and have it be accurate?

 

[picklemogg]

Assuming you are using local node inclusion if queing at local site as well then yes. If you for example queue calls to the skill locally first and then wait 10 seconds prior to queing to net sites there will be a 10 second period (while only queued local) where if the customer disconnects the abandon will only be counted against skill at local site. As long as skill is being queued at the same time at all sites this number would be accurate. We have done this exact thing for certain skillsets. Good luck.