Genesys CTI User Forum
Genesys CTI User Forum => Genesys CTI Technical Discussion => Topic started by: halflifecrysis on July 10, 2021, 02:57:49 PM
-
We are coming from inContact to Genesys. With inContact ASA was once the call was routed to the agent, how long did it take the agent to answer the phone, our average was 12 seconds for the dept. With Pure, it says our ASA is 22-40 seconds, that did not just happen overnight. I want to know, like before, once it routes to you, how many seconds of the phone ringing is happening. Our agents phones are not ringing 40 seconds lol. I'd love to know what Pure is including in that time that I contact wasn't. **I'd also love to know for those of you using this what is your ASA goal?**
So anyway we have some newbie WFM folks trying to compare Pure ASA to our previous inContact ASA goal of 12 seconds, that obviously will need to change and they aren't computing what we are trying to tell them. Any ideas or thoughts? Can I even find what I deem real ASA?
Sent from my Pixel 3 XL using Tapatalk
-
Sounds like you need a statistic monitoring the average time, comparing the total callringing duration to the number of callanswered action, and monitoring that for a group of agents.
as to what you are monitoring - could you share the statistic definition for the metric(s) used; then I think we can help you define what you're actually measuring.
And if this wasn't an engage on prem question... I might be a little off the rails with the answer ;)
Skickat från min Mi MIX 3 5G via Tapatalk
-
Genesys is just a CTI, when and how the call reaches your agents depend on that pbx or sip server (if using Genesys SIP server).
Usually the answer would be immediately, but have seen scenarios where network issues at agent side can make that not to happen.
To be sure you would need to follow by logs one call you mark as delayed and see what is going on. Possible cause could be trying to reach agent A, no network at his side and TTL for sip is still marked as ok, then URS and TServer will try to route to agent B but only after a few seconds trying on agent A. So... The guys who are implementing this Genesys need to understand your scenarios and then do proper configurations to try to reach the best possible scenario. Also you will need to try to fix possible agent side issues or network issues.
But without logs I am just sharing some experience, of course, there are many other possibilities too like high cpu usage for example which is HW related at server side...
So again, without logs, hard to say
Enviado de meu SM-N9600 usando o Tapatalk
-
With inContact the definition of ASA is the average amount of time to answer the call after the client (contact) chose the option for that agent (or their queue). So client presses 1 for payment, the system says I'm sending the call to Sarah, then it starts a countdown. Sarah sometimes answers within 10 to 12 seconds, on her end that may mean her phone rang three times.
Purecloud/Geneys says ASA is the time an interaction waits in queue before an agent answers it and denotes a calculation in it's definition of Total time to answer / Number answered.
With inConta I think our voice was carried by Sprint. With Pure, there are no other carriers involved that I know of, they handle it all.
But ASA in Pure is looking like 22-30 seconds vs. 10 or 12 seconds with inContact.
No one can really explain it, but I can say the overnight we didn't become twice as slow to answer. I haven't found anyone who manages another call center to tell me what their numbers look like so I have a baseline.
Sent from my Pixel 3 XL using Tapatalk
-
Well any good Genesys engineer would be able to explain the life cycle since second 0 up to when the call arrives to the agent.
Without knowing your architecture/implementation, it is hard to say.
Maybe you are comparing 2 different things that sound similar to you but are not exactly.
Carrier has nothing to do.
-
Try creating a new statistic in statserver (a section in statserver configuration, 'AverageCallAnsweredTime' or any name you'd prefer) with the following options set:
Subject: DNAction
Category: AverageTime
MainMask: CallRinging
RelMask: CallAnswered
Objects: Agent,GroupAgents
This will give you the average time from offered to an agent until agent answers, for the agent(s) the statserver client is requested to monitor.
There are some things to consider, like:
A call that has been offered but not accepted by agent for whatever reason will impact the average, since the ringing duration will be collected regardless of the final outcome.
This is one example, there are probably more, depending on the line of business and N other factors.
You'll probably find tweaks you want to do to disregard some scenarios by being more precise about what actions you define in the relative mask, and possibly MainMask as well.
As I said before, if you could share the definition you're using currently with the unexpected results we could help with a description of what you're monitoring.
And, if you want to define statistics you should have an experienced genesys engineer onboard, someone familiar with statserver configuration, like cavagnaro suggested.
It will be a lot of trial and error otherwise.
Br
Skickat från min Mi MIX 3 5G via Tapatalk
-
You're going to have to point me in a direction. At my level of my org we don't get much support.
I understand PureCloud can have plugins, is Star Server a plug in or part of the included package of PureCloud?
As it stands we have basic reporting, quality management forms for reps, WFM got turned on recently. So we are taking baby steps. We went from being experts in inContact at our org level to having it taken away and told we had to use corp software (Pure) so our HQ was not paying for 2 platforms. So we went from having all the keys, contacts to being pushed down an org level without much support.
So I've shared to someone the comments made here in stat server, but we still haven't heard that term hence my questions above.
I appreciate your patience with my Q.
Sent from my Pixel 3 XL using Tapatalk
-
Wait... What Genesys do you have? Genesys Cloud? Pure Cloud? Can you show a screen of agent desktop?
There are now up to 3 versions of Genesys... I was talking about Genesys Engage and think Genfrm was too...
Please clarify so we can help further
Enviado de meu SM-N9600 usando o Tapatalk
-
https://imgur.com/a/myzTwgB
Genesys Cloud, but everyone calls it Pure.
Here are the latest notes on the latest release pushed to us. https://help.mypurecloud.com/releasenote/july-14-2021/
@cavagnaro
-
Ah ok now gets more sense.
You have Edge servers right?
Can you get a detail call flow report for one problematic call?
Enviado de meu SM-N9600 usando o Tapatalk
-
I have looked under all the reports I have access to, I don't see that report.
https://docs.genesys.com/Documentation/GCXI/9.0.0/User/HRCXIIxnHndlngAttmpt
I researched what it is, but I have nothing titled interaction handling report.
Are you saying this stat server is NOT something we need to look into?
-
Yes, statserver and that metric is for Engage.
Even the logs I mentioned was for Engage.
What interaction reports you have available? Interaction Flow Report? Do you have it under Details report?
Enviado de meu SM-N9600 usando o Tapatalk
-
Ok, so genesys cloud :)
It sounds to me like you need to look at this documentation:
https://help.mypurecloud.com/articles/new-agents-performance-summary-view/
As i understand your requirement this is an agent performance metric, the 'reaction' time for the agent once the phone is ringing until the call was picked up.
From the link above I'd say you need to divide 'Total Alert' by 'Answered' to get that value.
There does not seem to be such a metric already calculated in the reports, at least not at first glance.
You could probably benefit from exploring https://help.mypurecloud.com/articles/about-reports-views-and-dashboards/
Sorry about the engage-stuff, i thought you had that flavor...
About ASA: that is a queue-metric so that would be reflecting things like your staffing, both the time the call is waiting in queue and ringing at an agents desktop - the full duration from queued to accepted.
https://help.mypurecloud.com/articles/queue-performance-detail-view/
Br
-
Gen_frm a better approach is to measure the whole path, starting with the IVR, because there could be for example a WS in the path hidden from customer that acts as a bottleneck. Measuring just the agent won't solve his main question, why the call since customer press an option in the IVR to the agent has such delay. You are talking only after IVR and queue involved, removing those from the equation is not a good idea if he wants to understand the delay as a whole.
See that he is a new user and maybe his concepts from other platform are not quite the same in Genesys, therefore better to provide a complete look.
Enviado de meu SM-N9600 usando o Tapatalk
-
I thought the question was about metric definitions, I didn't think that OP actually thought there was a problem.
I agree with you that for troubleshooting you need from second 0 and onwards.
Br
Skickat från min Mi MIX 3 5G via Tapatalk
-
[quote author=cavagnaro link=topic=12061.msg54129#msg54129 date=1626277686]
Gen_frm a better approach is to measure the whole path, starting with the IVR, because there could be for example a WS in the path hidden from customer that acts as a bottleneck. Measuring just the agent won't solve his main question, why the call since customer press an option in the IVR to the agent has such delay. You are talking only after IVR and queue involved, removing those from the equation is not a good idea if he wants to understand the delay as a whole.
See that he is a new user and maybe his concepts from other platform are not quite the same in Genesys, therefore better to provide a complete look.
Enviado de meu SM-N9600 usando o Tapatalk
[/quote]
Totally believe the whole path is important, especially for the big picture, but my current WFM team is reporting that whole path as a performance metric of the team and they have no knowledge of what other interruptions could cause those delays. So they are IMHO speaking from a place they shouldn't.
If they want to tie this metric back to agents, as we were able to do that in the past with inContact and this new WFM team had our old stats, then they need to be able to separate the two stats to under big picture and speed of answer from the agent once the call is routed and ringing to them. That is where I am at.
Just trying to learn more so I can speak better to it.
-
[quote author=gen_rtfm link=topic=12061.msg54128#msg54128 date=1626277414]
Ok, so genesys cloud :)
It sounds to me like you need to look at this documentation:
https://help.mypurecloud.com/articles/new-agents-performance-summary-view/
As i understand your requirement this is an agent performance metric, the 'reaction' time for the agent once the phone is ringing until the call was picked up.
From the link above I'd say you need to divide 'Total Alert' by 'Answered' to get that value.
There does not seem to be such a metric already calculated in the reports, at least not at first glance.
You could probably benefit from exploring https://help.mypurecloud.com/articles/about-reports-views-and-dashboards/
Sorry about the engage-stuff, i thought you had that flavor...
About ASA: that is a queue-metric so that would be reflecting things like your staffing, both the time the call is waiting in queue and ringing at an agents desktop - the full duration from queued to accepted.
https://help.mypurecloud.com/articles/queue-performance-detail-view/
Br
[/quote]
Right now, my alert number is a static number like 41 and an agent my have an answer of 38. How to I or where else should I look to tie this to a seconds metric. Like a 17 second alert and a 12 second answer metric to divide?
-
I still would recommend to check the whole call flow to understand the times. Then we can suggest metrics, but without knowing what exactly is happening there, we might end up shooting to the wrong place.
With 1 call flow with that delay you can show us and we could provide better direction
Enviado de meu SM-N9600 usando o Tapatalk
-
[quote author=cavagnaro link=topic=12061.msg54127#msg54127 date=1626273840]
Yes, statserver and that metric is for Engage.
Even the logs I mentioned was for Engage.
What interaction reports you have available? Interaction Flow Report? Do you have it under Details report?
Enviado de meu SM-N9600 usando o Tapatalk
[/quote]
Under detail reports, Agent Quality, User Status, Interaction Detail, Login/Logout Detail.
Looking at Interaction detail, it gives us a Queue wait in seconds per interaction, but not clearly agent answer speed once it queued to them.
-
[quote author=halflifecrysis link=topic=12061.msg54132#msg54132 date=1626280829]
[quote author=gen_rtfm link=topic=12061.msg54128#msg54128 date=1626277414]
Ok, so genesys cloud :)
It sounds to me like you need to look at this documentation:
https://help.mypurecloud.com/articles/new-agents-performance-summary-view/
As i understand your requirement this is an agent performance metric, the 'reaction' time for the agent once the phone is ringing until the call was picked up.
From the link above I'd say you need to divide 'Total Alert' by 'Answered' to get that value.
There does not seem to be such a metric already calculated in the reports, at least not at first glance.
You could probably benefit from exploring https://help.mypurecloud.com/articles/about-reports-views-and-dashboards/
Sorry about the engage-stuff, i thought you had that flavor...
About ASA: that is a queue-metric so that would be reflecting things like your staffing, both the time the call is waiting in queue and ringing at an agents desktop - the full duration from queued to accepted.
https://help.mypurecloud.com/articles/queue-performance-detail-view/
Br
[/quote]
Right now, my alert number is a static number like 41 and an agent my have an answer of 38. How to I or where else should I look to tie this to a seconds metric. Like a 17 second alert and a 12 second answer metric to divide?
[/quote]The alert metric is defined as "The number of times agents receive an alert for interactions".
Check the column/metric named "Total Alert", defined as "The total time Genesys Cloud alerts agents before the agent accepts or declines an interaction". That should be the duration/seconds, the examples you've given looks like a count.
Br
Skickat från min Mi MIX 3 5G via Tapatalk
-
[quote author=gen_rtfm link=topic=12061.msg54135#msg54135 date=1626285053]
[quote author=halflifecrysis link=topic=12061.msg54132#msg54132 date=1626280829]
[quote author=gen_rtfm link=topic=12061.msg54128#msg54128 date=1626277414]
Ok, so genesys cloud :)
It sounds to me like you need to look at this documentation:
https://help.mypurecloud.com/articles/new-agents-performance-summary-view/
As i understand your requirement this is an agent performance metric, the 'reaction' time for the agent once the phone is ringing until the call was picked up.
From the link above I'd say you need to divide 'Total Alert' by 'Answered' to get that value.
There does not seem to be such a metric already calculated in the reports, at least not at first glance.
You could probably benefit from exploring https://help.mypurecloud.com/articles/about-reports-views-and-dashboards/
Sorry about the engage-stuff, i thought you had that flavor...
About ASA: that is a queue-metric so that would be reflecting things like your staffing, both the time the call is waiting in queue and ringing at an agents desktop - the full duration from queued to accepted.
https://help.mypurecloud.com/articles/queue-performance-detail-view/
Br
[/quote]
Right now, my alert number is a static number like 41 and an agent my have an answer of 38. How to I or where else should I look to tie this to a seconds metric. Like a 17 second alert and a 12 second answer metric to divide?
[/quote]The alert metric is defined as "The number of times agents receive an alert for interactions".
Check the column/metric named "Total Alert", defined as "The total time Genesys Cloud alerts agents before the agent accepts or declines an interaction". That should be the duration/seconds, the examples you've given looks like a count.
Br
Skickat från min Mi MIX 3 5G via Tapatalk
[/quote]
Okay, thank you - I did not scroll down past the (A's) to find the word alert tied to another phrase like "total alert" - Okay, I do have a total alert and it does display in seconds.
Do I take that total time and divide it by total alert? Meaning 2m 36 sec total alert for someone who had 23 alerts, are we theoretically saying that person has 7 seconds answer time to the alert?
And if someone has 16 alerts and 4m 29sec total alert, that person had a 15.8 (ASA)?
-
Yes, yet I think you should open a support ticket with Genesys to try to them to explain why that long time, also try to get the max min values
-
Something like that, yes.
For an AverageSpeedofAnswer statistic I would divide the 'Total Alert' duration by the number of Answers, not alerts.
If you use Alert and Total Alert you would get average alerting duration regardless of if the interaction offer was answered.
Br
Skickat från min Mi MIX 3 5G via Tapatalk
-
[quote author=gen_rtfm link=topic=12061.msg54138#msg54138 date=1626287651]
Something like that, yes.
For an AverageSpeedofAnswer statistic I would divide the 'Total Alert' duration by the number of Answers, not alerts.
If you use Alert and Total Alert you would get average alerting duration regardless of if the interaction offer was answered.
Br
Skickat från min Mi MIX 3 5G via Tapatalk
[/quote]
Guys we are coming closer to the answer here.... Thank you
If what you are saying is what I am looking for then my team has a speed of answer of 6.7 seconds, this is closer to the 8-12 range we had with inContact than the 20-40 seconds our WFM team is trying to say is happening. I of course will try to verify this. Thank you
-
[quote author=halflifecrysis link=topic=12061.msg54139#msg54139 date=1626287739]
[quote author=gen_rtfm link=topic=12061.msg54138#msg54138 date=1626287651]
Something like that, yes.
For an AverageSpeedofAnswer statistic I would divide the 'Total Alert' duration by the number of Answers, not alerts.
If you use Alert and Total Alert you would get average alerting duration regardless of if the interaction offer was answered.
Br
Skickat från min Mi MIX 3 5G via Tapatalk
[/quote]
Guys we are coming closer to the answer here.... Thank you
If what you are saying is what I am looking for then my team has a speed of answer of 6.7 seconds, this is closer to the 8-12 range we had with inContact than the 20-40 seconds our WFM team is trying to say is happening. I of course will try to verify this. Thank you
[/quote]Best of luck to you!
Skickat från min Mi MIX 3 5G via Tapatalk