CRM includes Sentiment Analysis and Conversation Summary data for calls. Plus, you can view transcripts on call detail screens.
Here’s an example of how it looks in the Unprocessed Call List:
What is Sentiment Analysis?
We use an AI model to evaluate each call transcript and assign a sentiment score: Positive, Neutral, or Negative. This score is based on the total average sentiment of the entire call transcript.
Sentiment Analysis helps you quickly identify at-risk customers or call outcomes without replaying full calls, and supports call coaching and quality assurance by surfacing sentiment trends by agent and location.
What is Conversation Summary?
This feature generates a concise AI summary (typically 1-3 sentences) for every call transcript, capturing the purpose, key outcomes, and next steps of the conversation.
Conversation Summary reduces time spent reviewing calls by letting you triage interactions directly from the history list and reports. Plus, this feature enhances visibility into call quality and outcomes across your entire team. This gives you extra value with CRM as an extra communication intelligence layer, beyond our regular call recording management.
Where can I find Sentiment Analysis and Conversation Summary information?
You’ll find this information readily available in the Interaction History menu option (both Unprocessed and All calls), while grading calls from the employee list and to-do list, on the Customer history tab, and the Lead history tab.
Where can I find in-app call transcripts?
You can find call transcripts on any call detail screen.
Note: All call transcripts automatically redact any sensitive Payment Card Information (PCI).
Are there any in-app reporting metrics related to Sentiment Analysis and Conversation Summary?
Yes, you can view this information in our Agent Sentiment Summary and Call Sentiment Report.
Where can I access Sentiment data via Snowflake?
If you use Snowflake data reporting, the sentiment data will be delivered via the Snowflake data share and can be found within the view called "Sentiment_Analysis.” This view includes the following fields:
Column Name |
Description |
summary_id |
Unique identifier for the AI-generated summary record. |
call_log_id |
Reference ID linking to the original call record. |
account_id |
Unique identifier for the client account. |
location_id |
Location where the call occurred. |
sentiment_score |
Numerical value representing overall call sentiment. |
sentiment_label |
Categorical result given (e.g., Positive, Neutral, Negative). |
summary |
A concise text recap of the conversation highlights. |
call_date |
The timestamp or date the call was placed. |
agent |
The staff member who fielded the call. |
recording_url |
A direct link to the stored audio file of the conversation. |