Workforce Engagement Management

We have a specific topic that Checks for Resolution on both the customer and agent side of the call, that is having great results but we are seeing some weird misses. we have one short phrase that should have been picked up on a recent call and was not. The phase was "that's all", and in the interaction that missed it the customer said "that's all I needed", so it should have triggered the topic. Any idea why it would not? The phrase was set with the "Topic Default" strictness, and the Topic default was Med-High, set for bpth Participants.

#SpeechandTextAnalytics

------------------------------
Conor Twomey
American Family Mutual Insurance Company, S.I.
------------------------------

Hi Conor,

The more phrases you provide, the more accurate it will be. So I would add all sorts of variations not just he base phrase. Eg Thanks that is all, that's all thanks, that's all I need, that is all I need, etc not just partial, base phrases.

This gives the AI match greater understanding of what to look out for, you don't need to listed every variation, the language modal will figure it out, but more is definitely better.

Depending on the accuracy of the transcript, Med-High could also be too high.

Speaking of which, the other thing to check is, in that call, what does the transcript show, does it show exactly what the customer said, or was there a transcription error / difference? 

I'm not expert on this, so just my two cents, hope it helps. 

------------------------------
Anton Vroon
------------------------------

Hi Conor, 

Have you been experiencing additional misses like this?

I suggest opening a SERVOPS ticket with additional details about the conversation, that way we can investigate the root cause of what happened. 

------------------------------
Adi Goodman
Genesys - Employees
------------------------------