Skip to main content

How to Configure the Topic Analysis Panel

Learn how to configure the Topic Analysis panel in the Loop Experience Platform using NLP sentiment clustering, tag taxonomies and AI to analyze open-ended feedback.

A
Written by Arutyun Airapetyan

Summary

This How-To Guide provides step-by-step instructions for platform administrators, location managers, and internal staff to build and configure the Topic Analysis matrix panel in Dashboard Projects. By combining natural language processing (NLP) sentiment scoring with predefined tag taxonomies, the Topic Analysis panel clusters unstructured, open-ended feedback comments into key operational themes to expose recurring root causes across physical sites and channels.

Prerequisites & Permissions

  • Primary Audience: Platform Administrators, Location Managers, Quality Assurance Officers, and Benbria Internal Staff.

  • Permissions Required:

    • Account Owners: Full permissions to add, configure, move, resize, and remove Topic Analysis panels across global or regional Dashboard Projects.

    • Location Admins: Permissions to view, filter, drill down into, and export topic insights for Dashboard Projects assigned to their physical locations.

Step-by-Step Instruction

Step 1: Open the Target Dashboard Canvas

  1. Log in to the Loop Experience Platform.

  2. From the main left navigation menu, click Projects and select Dashboard Projects.

  3. Click the name of the dashboard project you wish to configure (e.g., Operational Performance Dashboard or Executive Quality Overview).

  4. Click Edit Dashboard in the upper right header to unlock the editing canvas.

Step 1: Setup the Topics

  1. Contact your Customer Success Manager to request this feature and to align on the set of topics to be analyzed. Topics are configured per dashboard and are focus on a single Category.

Step 2: Add the Topic Analysis Panel

  1. Click + Add Panel or + Create Custom Panel on the toolbar and select Topic Analysis.

  2. Under the General tab, enter a clear Title (e.g., Topics & Sentiment Trend) and optional Description.

  3. Unser the Metric tab, select the desired Category such as Comments.

  4. Click Save Panel.

💡 TIP: Review your organization's keyword dictionaries in Admin / Settings > Tags & Keywords quarterly to add local slang or common typos, ensuring maximum accuracy for NLP topic clustering.
​

⚠️ NOTE: Sentiment classification relies on text analysis; comments without open-ended text or matching keywords will be excluded from topic counts, but remain accessible in full Result Table panels.
​

Frequently Asked Questions (FAQs)

Q: Can I click directly on a topic bar to see the individual comments?

A: Yes. Clicking any topic row or sentiment segment (e.g., negative sentiment under Facilities) opens the Underlying Data Drawer, displaying the exact survey comments and associated tickets that generated that score.

Q: How does the panel handle comments that match multiple tag categories?

A: If a single respondent comment contains keywords belonging to multiple topics (e.g., mentioning both "cold food" and "rude staff"), the platform increments the volume count for both categories while preserving the overall submission record.

Q: Does modifying tag keywords in Admin / Settings update past Topic Analysis panels?

A: Auto-tagging rules apply to feedback upon submission. Updating keyword dictionaries refines topic classification for all future incoming submissions without altering historical records.

Did this answer your question?