|
本帖最后由 Apuroos073 于 2023-3-28 12:41 编辑
Platform from machine learning That is, we tell the software what constitutes a valid connection, and based on these parameters, it will indicate which ones meet this criteria. By default, the criteria that lead to low call quality are: time waiting for customer service; quiet time during the call (customer waiting too long for a resolution - this could mean); crosstalk - agent and customer connecting Talking at the same time during the call, which may indicate communication noise; Short call duration for example, a 30-second call may score lower because.
The bot knows it is impossible to solve so many problems in such a short call); Call Analysis - that resources understand the stage of contact and ensure the confidence of lebanon phone number list the team. After all, it will be possible to understand whether arguments are appropriate and how consumers react to them. This way, the mistakes and successes of the chat can be identified so that it can be optimized later. Understanding other basic data for analyzing calls Call Score is critical to assessing the most effective calls.

However there are other data that can be observed and can collaborate to make this diagnosis. Average service time is one of them. This indicator tells you the duration of the call. Hence, it helps the pro to know if it is too short for the conversion or more than expected. It's also interesting to use it in support cases to see how many minutes it takes to resolve a certain issue. Another similar statistic is customer retention. It indicates the number of people who continue to talk until the end of the contact instead.
|
|