Train Verbatim

The Train Verbatim section allows users to review, categorize, and train AI models using actual customer feedback responses (verbatims). It enables you to assign tags, adjust sentiment, and refine text analysis results, ensuring that the system accurately understands customer feedback themes.

By manually reviewing and tagging responses, users help the AI improve tagging accuracy, sentiment detection, and topic classification for future analyses.

Key Components

Verbatim Responses

This section displays individual customer feedback responses exactly as submitted by respondents.

Each verbatim typically includes:

• Respondent name

• Feedback text

• Response date

• Detected sentiment

• Assigned tags

This allows analysts to read the full context behind customer opinions and validate automated insights.

Sentiment Selection (Manual Sentiment Training)

Each response is assigned a sentiment classification which can be manually adjusted.

Available sentiment options:

• Positive – Customer expresses satisfaction or praise

• Neutral – Customer feedback is informational or balanced

• Negative – Customer expresses dissatisfaction or complaints

Users can change the sentiment manually to improve the accuracy of AI sentiment analysis over time.

Tagging Verbatims

Tags help categorize feedback into specific themes or topics.

Users can:

• Assign tags to a verbatim

• Remove existing tags

• Add multiple tags to a single response

This helps organize customer feedback into structured insights such as:

• Delivery and Shipping Experience

• Device Condition and Quality

• Customer Service and Support

• Pricing Transparency

Tagging responses improves the accuracy of Tag Analysis and trend tracking.

Manage Tags

The Manage Tags panel displays the list of all available tags along with the number of responses associated with each tag.

Key capabilities include:

• Search tags using the search bar

• View tag usage counts

• Select a specific tag to manage or edit

Selecting a tag allows you to perform additional tag management actions.

Add Tags

Users can create or import tags using the Add Tags option.

Available methods include:

Generate AI Tags

Automatically generates tags based on patterns detected in the verbatim responses.

This helps quickly create meaningful categories from large volumes of feedback.

Add Tags Manually

Allows users to create custom tags manually.

This is useful when organizations want to track specific business themes or categories.

Example:

• Refund Delay

• Product Authenticity

• Packaging Quality

Import from Survey

Imports tags directly from survey questions or response options.

This ensures alignment between survey structure and feedback analysis.

Import from CSV

Allows users to bulk upload tags using a CSV file, making it easier to manage large tag libraries.

Tag Management Options

When a single tag is selected, additional tag management actions become available.

These include:

Generate Subtags with AI

Automatically generates subcategories within a tag based on the feedback context.

Example:

Parent Tag: Delivery Experience
Possible Subtags:

• Late Delivery

• Fast Delivery

• Packaging Issues

Create Subtags Manually

Users can manually create subcategories within a tag to provide deeper analysis.

Add to Group

Allows tags to be organized into groups for better categorization and reporting.

Convert to Subtag

Converts an existing tag into a subtag under another main tag.

This helps create structured tag hierarchies.

Edit Tag

Allows users to rename or modify the tag.

Delete Tag

Removes the tag from the system.

Note: Deleting a tag may affect previously tagged responses.

Tag and Train Verbatim Panel

This panel shows the verbatim responses along with sentiment and applied tags.

Users can:

• Assign tags to a response

• Remove tags

• Add additional tags

• Adjust sentiment classification

These actions help train the AI model to better understand feedback context and sentiment.

Search Verbatim

The Search Verbatim option allows users to quickly find specific responses using keywords.

This is useful for:

• Investigating particular issues

• Reviewing feedback related to a specific topic

• Validating tag assignments

Run Analysis

The Run Analysis option allows users to apply automated text analysis to the verbatim responses.

Available analysis options include:

AI Tag Analysis

Automatically detects and assigns tags to responses based on text patterns.

AI Sentiment Analysis

Automatically determines the sentiment of responses (Positive, Neutral, Negative).

Keyword Analysis

Identifies important keywords and recurring phrases within the feedback dataset.

This helps uncover hidden themes and emerging topics.

Filtering Options

Users can filter verbatim responses based on:

• Survey Question – Example: “What’s the primary reason for your score?”

• Tags – Filter responses by specific themes

• Sentiment – View only positive, neutral, or negative feedback

These filters help users focus on specific areas of feedback for deeper analysis.