cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
441
Views
2
Helpful
3
Replies

Abandoned Call Back

mightyking
Level 6
Level 6

Hello Experts,

We offer to our clients the option to chose to be called back instead of waiting in queues.

We receive an excessif number of abandoned calls in the CUIC report regarding the Callback calls. I am unable to find a reason for this huge number. As you can see we received 136 calls and 88 of them were abandoned. Could someone please help me to understand if I am missing anything?

 

mightyking_1-1720527652879.png

Thanks,

MK

 

 

 

2 Accepted Solutions

Accepted Solutions

Jonathan Schulenberg
Hall of Fame
Hall of Fame

CCX does not have a native queued callback feature. There are several IVR scripts floating around that attempt to [crudely] emulate that experience but none that I have encountered are perfect. Every incoming call to an IVR script is assumed abandoned until either: 1) the Select Resource step successfully connected the contact/call to an resource/agent; or, 2) the contact/call is explicitly marked as handled within the IVR script using the Set Contact Info step. I don't know the details of your callback IVR script(s) but the most likely explanation is that the abandons represent failed/lost callback requests.

One of many possible causes is the script queued loop ran until it hit the maximum step count limit (default: 1000) and crashed. You can try to fortify the script logic against some failure scenarios such as the step count limit but not all of them. For example, if the CCX cluster fails over the callback request is just gone.

Webex Contact Center has a native implementation of this feature, if migrating is an option. If you're sticking with CCX the recommended approach is to use a direct preview outbound campaign instead so the callback requests are actually stored in the database. The biggest challenge with that approach is that campaign calls are only offered to agents sitting in a ready state (i.e. there aren't any incoming calls); you can't interleve them between inbound calls sitting in queue.

View solution in original post

Sure it's possible. It just depends how fast the script is running through steps. For example, 30 seconds of hold vs. 30 minutes is going to have a major impact on how fast the queued loop iterates.

There are dozens of possible causes though, this was just an easy example to cite. If I were troubleshooting this I'd copy the script and configure it to use a test CSQ that you're the only agent skilled in. Just sit there in a not ready state and let the CRS Editor run in a reactive debug mode for at least as long as the longest real callback request waits. If it crashes, think very carefully about what was happening before you click OK on the error dialog. IIRC all the variables will empty and the reactive debug session ends the instant you acknowledge that error.

View solution in original post

3 Replies 3

Jonathan Schulenberg
Hall of Fame
Hall of Fame

CCX does not have a native queued callback feature. There are several IVR scripts floating around that attempt to [crudely] emulate that experience but none that I have encountered are perfect. Every incoming call to an IVR script is assumed abandoned until either: 1) the Select Resource step successfully connected the contact/call to an resource/agent; or, 2) the contact/call is explicitly marked as handled within the IVR script using the Set Contact Info step. I don't know the details of your callback IVR script(s) but the most likely explanation is that the abandons represent failed/lost callback requests.

One of many possible causes is the script queued loop ran until it hit the maximum step count limit (default: 1000) and crashed. You can try to fortify the script logic against some failure scenarios such as the step count limit but not all of them. For example, if the CCX cluster fails over the callback request is just gone.

Webex Contact Center has a native implementation of this feature, if migrating is an option. If you're sticking with CCX the recommended approach is to use a direct preview outbound campaign instead so the callback requests are actually stored in the database. The biggest challenge with that approach is that campaign calls are only offered to agents sitting in a ready state (i.e. there aren't any incoming calls); you can't interleve them between inbound calls sitting in queue.

Thank you Jonathan,

The Max Number of Executed Steps in our cluster is set to 2000. Do you still think this could be one of the reasons for the 88 abandoned calls out of 138 calls?

Thanks,

MK

Sure it's possible. It just depends how fast the script is running through steps. For example, 30 seconds of hold vs. 30 minutes is going to have a major impact on how fast the queued loop iterates.

There are dozens of possible causes though, this was just an easy example to cite. If I were troubleshooting this I'd copy the script and configure it to use a test CSQ that you're the only agent skilled in. Just sit there in a not ready state and let the CRS Editor run in a reactive debug mode for at least as long as the longest real callback request waits. If it crashes, think very carefully about what was happening before you click OK on the error dialog. IIRC all the variables will empty and the reactive debug session ends the instant you acknowledge that error.