Sentiment analysis leverages AI-powered prediction to understand customers' expectations and prioritize better by predicting real-time conversation sentiment for the latest customer message. This process offers insights into customer emotions, enabling real-time monitoring and timely interventions to improve customer experiences.


This article contains:

Benefits of Sentiment Analysis

  • Enhanced Customer Satisfaction: Address sentiments in real-time to improve overall customer experience.
  • Chat Prioritization: Categorize and prioritize conversations based on sentiment to ensure prompt resolution.
  • Churn Reduction: Identify and proactively engage at-risk customers to reduce churn.
  • Escalation Prevention: Quickly identify and address potential issues to prevent escalations.
  • Manage Negative Sentiments: Customize sentiment score ranges for precise categorization, effectively handling large volumes of negative sentiments.
  • Efficient Automations: Utilize automations to detect and respond to changing sentiments swiftly.
Note: Changes to sentiment score ranges and settings will apply to all conversations, both new and existing.

Understand Sentiment Scores

Sentiment Score

Sentiment scores provide a quantifiable measure of the sentiment expressed within the text, utilizing a predictive model that assigns scores to reflect sentiment intensity. Lower scores indicate negative sentiments, while higher scores reflect positive sentiments. This allows for an understanding of the emotional tone of the customer's message.

Sentiment Score Ranges

Sentiment scores range from 0 to 100, with 0 indicating the most negative and 100 indicating the most positive. By default, the scores are classified as follows:

  • 10 to 30: Negative
  • 31 to 70: Neutral
  • 71 to 100: Positive

Customization Options

Customization options to categorize Sentiment score ranges into negative, neutral, or positive based on business requirements are available. This option will help manage the large volume of negative sentiments often seen in customer conversations. After you set your sentiment score ranges, all new conversations created from that moment on will reflect the new ranges. 

Set Up Sentiment Analysis

Enable or Disable Sentiment Analysis

  1. Go to Admin settings.
  2. Click the Freddy icon to access the list of available AI features.
  3. Navigate to the Sentiment Analysis feature.
  4. Toggle the switch to the right to enable the feature or to the left to disable it.

Configure Advanced Settings

  1. Click on Configure next to the toggle beside the Sentiment analysis feature.
  2. Under the Set sentiment score range section, drag the slider on the sentiment scale to customize ranges for negative, neutral, or positive in increments of 5.
  3. Click on Reset to Default if you want to return to default ranges.
  4. Click Save to apply the changes.

Use Sentiment Analysis

  1. Login to your account and go to Conversation Inbox.
  2. Access the conversation list: Each conversation will be tagged with its respective sentiment. Hover over the sentiment to view the sentiment score.
  3. Prioritize based on sentiment: Consider the sentiment for each conversation, especially if the sentiment is negative and the score is higher.
  4. Filter/Sort by sentiment: Filter conversations by their sentiment - positive, negative and neutral.
    Sort conversations by sentiment to address urgent concerns.

  5. Address high negative sentiment score and high priority: Focus on conversations with both high negative sentiment score and high priority to address customer concerns promptly.
  6. Provide updates and monitor sentiment changes: Observe changes in sentiment after providing an update or resolution.

By following these steps, agents can effectively utilize Sentiment Analysis to prioritize and address customer sentiment, improving customer experiences and issue resolution.

The enablement of optional functionality is subject to certain feature-specific terms and conditions set forth in the Freshworks Supplemental Terms.