Queue Statistics.

Answered Question
Sep 2nd, 2008

Hi,

I intend to fetch the queue statistics from the ICM AW/HDS database. I am looking for statistics such as, Total Calls in queue, Ababndoned calls, Calls answered within service level, calls that crossed the service level threshold etc.

I have identified one table, Call_Type_Real_Time table that will give me all this info.

But I am not convinced with the idea that this should be the table I should be looking for.

I am looking for the statistics of the calls that are queued for a specific Skill group.

The Call type is associate with a script. This script may have multiple Queue to skill group nodes. The calls may be queued at each skill group. So when this happens, the Call_Type_Real_Time table will give me, say a total queue time, which is an addition of the times the call has stayed in queue for each skill group.

The other option is to use the Skill_Group_Real_Time table, but it does not give me info on the Abandoned calls, average delay etc, stats.

Can someone guide me on this?

Thank You,

-Bhaskar

I have this problem too.
0 votes
Correct Answer by cvenour about 8 years 3 months ago

Understanding Call Types and Skill Groups

My first step when developing an IPCC solution is to develop a call flow diagram, which shows call legs from each potential route start point (a DN, an Agent [xfers], a Route Point, etc), which Call Types and Scripts that leg can pass through, and which Skill Groups the call can end up on.

If you do this properly, you can put the diagram into the hands of the business or a technical person, and they can immediately understand the relationships between the different elements of the system.

-----------------

RE: Skill Group and Call Type stats

Generally speaking an ACD queues a call to a specific target resource group.

By comparison, IPCC can consider any number of target resource groups, simultaneously.

This means that looking at the Skill Group level will only provide you with part of the picture, because it only looks at one of a possible many possible target resource groups. Not to mention that the same Skill Group can be used in any number of scripts, so there is no way to know where a call is coming from.

As such, we have the concept of the Call Type as the meta-level data container, because it provides both source data (that is, it is the type of the call) and target data (it includes all of the skill groups under consideration).

-----------------

RE: Specific Skill Group stats

When a call hits a Queue to Skill Group node, the ICM router queues the request (not the call - as Geoff outlined above) against ALL of the skill group resources you have included in the QtoSG node.

What happens from a stats perspective is that the RouterCallsOffered count is incremented for ALL the Skill Groups under consideration.

When the router selects a target:

1. The CallsHandled count for the successful skill group is incremented, indicating that this skill group handled the call.

2. The RouterCallsDequeued count for each of the unsuccessful skill groups is incremented, indicating that this skill group did NOT handle the call.

-----------------

Hope that helps.

Regards,

Corey.

I wouldn't go that far.

A VDN in Avaya can issue a host-based route request (adjunct route) or control a ACDQ (a vector controlled ACDQ) through vector scripting, but the call type is just a classification method.

When a call arrives at the Call Router, it comes in on a dialed number. If the dialed number is associated with a call type, and the call type mapped to a routing script, it will execute that script.

So you have to have a call type or there is no routuing.

If all that script did was queue the call to the "HelpMe" skill group, then that single call type measures all the important parameters of the call; from the moment the call hit the ICM router until the agent is chosen; and since it is the only call type, will be recorded in the TCD record. Life is easy, and life is good.

But in a normal call center, call types are also used to count the segmentation of calls - as a result of a DB lookup, or a caller's response in the IVR.

As we classify the call, we change the call type to peg these counters. These call types almost always have "num. offerred" = "num. overflow out". No abandons or ASA.

Finally we get to the ICM routing script that has a number (possibly) of queue to skill group nodes, with some overflow characteristics and so on. Can be quite complex.

At the top of this script, we set a new call type - a "queuing call type".

It's no different to any other call type (though I always name these with a trailing "_Q" so they stand out in WebView); but because it is the last call type, the number that overflow out is 0.

This call type will measure the customer experience - ASA and service level, and the abandon count (not the time to abandon though; that is measured from the moment of entry to the Call Router).

Regards,

Geoff

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (2 ratings)
Loading.

Bhaskar, Haven't we been through this a number of times? I'm sure we have had this conversation before.

>But I am not convinced with the idea that this should be the table I should be looking for.

It is. How can I convince you?

>say a total queue time, which is an addition of the times the call has stayed in queue for each skill group

Not a sum. If you queue to multiple skill groups, the time in queue is the same for all and is recorded against the call type.

>other option is to use the Skill_Group_Real_Time table

As you say, this is NOT an option.

Regards,

Geoff

bhaskar27in Tue, 09/02/2008 - 08:56

Hi Geoff,

I am sorry about coming back again with the same issue. Its my ignorance with regard to Cisco.

I am trying to come to terms with Cisco's terminology here. I have so far worked with Nortel and Ericsson PBX's, where the use of a Queue is different from what I am trying to understand with regard to Cisco.

I appreciate your responses on this.

By the way, My understanding so far are as follows:

1. If a script has multiple queue to skill group nodes, the call is typically queued to just ONE queue.

2. The statistics for this queue can be fetched based upon the Call Type that is associated with the script.

My question here is related to the "Queue to Skill Group Node" (Which has been confusing me so far).

What exactly happens when this node is encountered during the script execution? Say when the prior node is Translation Route to VRU and the next node is Run External script.

Thanks for your patience and awaiting your response.

-Bhaskar

Indeed - the TDM environment is quite different. You have to throw out some of your old concepts.

>1. If a script has multiple queue to skill group nodes, the call is typically queued to just ONE queue.

No, it is in all skill groups.

Your problem is thinking that it is queued there. It's not really queued in the skill group. The call is queued somewhere (CVP, IP IVR) and the skill group is a set of logical targets.

Regards,

Geoff

Correct Answer

I wouldn't go that far.

A VDN in Avaya can issue a host-based route request (adjunct route) or control a ACDQ (a vector controlled ACDQ) through vector scripting, but the call type is just a classification method.

When a call arrives at the Call Router, it comes in on a dialed number. If the dialed number is associated with a call type, and the call type mapped to a routing script, it will execute that script.

So you have to have a call type or there is no routuing.

If all that script did was queue the call to the "HelpMe" skill group, then that single call type measures all the important parameters of the call; from the moment the call hit the ICM router until the agent is chosen; and since it is the only call type, will be recorded in the TCD record. Life is easy, and life is good.

But in a normal call center, call types are also used to count the segmentation of calls - as a result of a DB lookup, or a caller's response in the IVR.

As we classify the call, we change the call type to peg these counters. These call types almost always have "num. offerred" = "num. overflow out". No abandons or ASA.

Finally we get to the ICM routing script that has a number (possibly) of queue to skill group nodes, with some overflow characteristics and so on. Can be quite complex.

At the top of this script, we set a new call type - a "queuing call type".

It's no different to any other call type (though I always name these with a trailing "_Q" so they stand out in WebView); but because it is the last call type, the number that overflow out is 0.

This call type will measure the customer experience - ASA and service level, and the abandon count (not the time to abandon though; that is measured from the moment of entry to the Call Router).

Regards,

Geoff

Correct Answer
cvenour Tue, 09/02/2008 - 14:24

Understanding Call Types and Skill Groups

My first step when developing an IPCC solution is to develop a call flow diagram, which shows call legs from each potential route start point (a DN, an Agent [xfers], a Route Point, etc), which Call Types and Scripts that leg can pass through, and which Skill Groups the call can end up on.

If you do this properly, you can put the diagram into the hands of the business or a technical person, and they can immediately understand the relationships between the different elements of the system.

-----------------

RE: Skill Group and Call Type stats

Generally speaking an ACD queues a call to a specific target resource group.

By comparison, IPCC can consider any number of target resource groups, simultaneously.

This means that looking at the Skill Group level will only provide you with part of the picture, because it only looks at one of a possible many possible target resource groups. Not to mention that the same Skill Group can be used in any number of scripts, so there is no way to know where a call is coming from.

As such, we have the concept of the Call Type as the meta-level data container, because it provides both source data (that is, it is the type of the call) and target data (it includes all of the skill groups under consideration).

-----------------

RE: Specific Skill Group stats

When a call hits a Queue to Skill Group node, the ICM router queues the request (not the call - as Geoff outlined above) against ALL of the skill group resources you have included in the QtoSG node.

What happens from a stats perspective is that the RouterCallsOffered count is incremented for ALL the Skill Groups under consideration.

When the router selects a target:

1. The CallsHandled count for the successful skill group is incremented, indicating that this skill group handled the call.

2. The RouterCallsDequeued count for each of the unsuccessful skill groups is incremented, indicating that this skill group did NOT handle the call.

-----------------

Hope that helps.

Regards,

Corey.

bhaskar27in Wed, 09/03/2008 - 21:30

Geoff and Corey,

Thank You for your valuable inputs and your participation in this discussion. I appreciate your responses which has educated me immensely on this topic.

You guys ROCK and thanks to Cisco for providing this wonderful arena.

Cheers!

-Bhaskar

Actions

This Discussion