Negative Sentiments or Needs Attention Alerts

The reason that we select only inbound customer comments is to reduce the number of alerts that you receive.  In our experience, Support Managers typically want to be alerted based on what their customers are saying. 

Here is an example of an alert for Negative Sentiment and Needs Attention Signals detected as per the suggested scenario above.  You can include the severity and customer name in the alert payload for additional context if needed.

_________________________

______________________________________________

 Sentiment Score or Needs Attention Score Alerts

Another alert that we recommend creating is when a case owned by a member of your team has a Sentiment or Needs Attention Score that meets a certain threshold.   The scores take into consideration the NLP signals but also response time.  I would recommend the following thresholds to start:

Needs Attention Score > 70

Sentiment Score <40

See examples below

______________________________________________

You can raise or lower these depending on your specific needs.  If you want the alerts to be more sensitive, you might reduce the threshold for Needs Attention Score and increase the threshold for Sentiment Score. 

______________________________________________

Likely to Escalate Alert

Another alert that we recommend creating is when a case owned by a member of your team is predicted as Likely to Escalate.  SupportLogic’s ML is constantly analyzing your cases for risk factors.  By setting up an alert for this scenario, you can proactively notify that a case is at risk even when you aren’t performing escalation prevention workflows in the UI.

______________________________________________

For this alert, we recommend configuring it per the example below

SupportLogic alerting capabilities are extremely robust and can meet a variety of use cases. We encourage you to experiment with them based on the needs of your organization.  

Office Hours All About Alerts Recording