Sentiment analysis shows the overall customer sentiment for a completed interaction. You can use it to identify negative customer experiences faster, prioritize interactions for review, add context to agent and team performance discussions, and track sentiment trends across completed interactions.

How sentiment scoring works

Sentiment analysis calculates a customer sentiment score between -100 and +100 for each completed interaction.

The sentiment result includes both the score and the classifier for the interaction.

The overall customer sentiment is classified using these thresholds:

Sentiment	Score Range	Description
Positive	> +46	The customer expressed satisfaction or a positive tone.
Neutral	-45 to +45	The customer's tone was professional or indifferent.
Negative	< -45	The customer expressed frustration or dissatisfaction.

The score is based on customer utterances in the interaction, with more weight given to utterances near the end of the interaction.

Before you begin

• Make sure your organization has the required AI Quality Management entitlement.
• Sentiment analysis must be enabled by your administrator in Control Hub.
• Real-time transcription must be enabled for the queue.
• The flow must be updated to support real-time transcription for the interactions you want to analyze.

Administrators can enable Sentiment analysis at the organization level or queue level in Control Hub. This lets organizations turn on the feature only for the queues where it is needed.

To enable Sentiment analysis in Control Hub:

1. Sign in to Control Hub and navigate to Services > Contact Center.
2. Navigate to Desktop Experience > AI Features in the navigation pane.
3. Turn on Sentiment analysis.
4. To enable the feature for selected queues, open the Queue level tab, select one or more queues, and enable the feature.
5. Make sure real-time transcription is enabled for those queues and that the corresponding flow is updated.

Supervisors can view customer sentiment in the Completed interactions table and in the interaction details view for a completed interaction. The sentiment score and classifier are displayed for each interaction.

In the Completed interactions table, supervisors can filter interactions by sentiment category.

Available filters include:

• Positive
• Neutral
• Negative

This makes it easier to find interactions that may need review, such as interactions with negative sentiment or interactions you want to compare across a team or queue.

This section answers frequently asked questions about using the Sentiment Analysis for your organization.

1. What does sentiment analysis measure?

   Sentiment analysis measures the overall customer sentiment for a completed interaction and classifies it as positive, neutral, or negative based on the sentiment score.

2. Is sentiment analysis enabled for the whole organization or by queue?

   Sentiment analysis can be enabled at the organization level or at the queue level in Control Hub. Administrators can select multiple queues and enable the feature for those queues.

3. What must be configured before sentiment analysis works?

   Real-time transcription must be enabled for the queue, and the flow must be updated to support real-time transcription for the interactions you want to analyze.

4. Where can supervisors view sentiment?

   Supervisors can view sentiment in the Completed interactions table and in the interaction details view for a completed interaction.