How to Manage EWT with a multiple priorities VQ

Architect & Flows
1

Hi,

I am trying to find some best practices to manage the following case:

  • I have multiple kind of customers that could call a CC (ex: High, Medium, Normal customers)
  • Each customer has different priorities
  • All customers call the same callflow.
  • All calls received on this callflow are answered by the same agent group (ex: say VAG1)
  • All calls are put in the same VQ (ex: says VQ1) and they are distributed to an agents based on their priority.

Now I need to implement EWT.
I need to announce the EWT (your EWT in less than, or is greater than …)

But how could I have an accurate EWT if I have multiple priorities in the same VQ

Do I have to create a VQ by priority?
Do I have to create a EWT statistic with a filter based on the priority?
Other solutions?

Thanks

Melvyn

2

you can’t if you have priorities for the same Queue

3

Break them on different target and then try to get the EWT from VQ associated to the target object.
Example

Target 1 is VIP_Ag1
Target 2 is Normal_Ag1

Each has its own VQ and times so you then inform or calculate based on those values. Not a single target object

4

Hi cavagnaro,

Thanks for your feedback.
I know that your solution could be the easy one and maybe the only one but I let you imagine how many VQ and AG I will have to manage if I have hundred and hundred of call flow.
And I could I have a accurate EWT if agents are part of multiple AG?

In this case the default EWT statistic available on URS could not be used

5

There is no easy solution for complex scenarios my friend, no magic clicks
As long as you configure the origDN correctly then yes, because EWT is calculated on agent group basis, not on agent alone

6

I might be wrong but

  • if different customers have different FIXED prioritities any predicition will be wrong - lower priority customers will never be answered at all (if there is more or less constant flow of higher priority calls).
  • as such probably function InVQWaitTime when single VQ is used will give prediction not much worse that any other method
  • and if you like to experiments you can try function RvqData which provide URS based statistics (including waiting time) based on queues that URS maintain internally
    (not on Virtual Queues). Might be a bit tricky as you have to make sure in such case that every category of cucteomrs sharing the same priority use their own
    router’s queue even though they might targets the same agent group.
7

You can use Increase Priority function to ensure that calls with lower priority will not stay endless in VQ.