How do I use spotlight to get root cause insights from my conversations?
In this article we'll be covering:
Summary of Spotlight
Permissions
Selecting Conversations
Running Spotlight
Viewing Evidence
Download conversation CSV
Summary of Spotlight
Spotlight helps you to uncover not just what is happening inside your conversations, but why. This AI-powered insight tool transforms how your teams can investigate customer issues. By analysing filtered conversation sets and surfacing the evidence behind each insight, Spotlight turns hours of manual analysis into seconds. Spotlight is your AI Business Analyst that automatically identifies patterns, issues, and opportunities across your customer conversations.
The result? A clear, AI-powered understanding of what customers are saying, why issues are happening, and where teams should focus next.
Permissions
Spotlight is available as part of our conversation intelligence suite. Contact your customer representative to enquire about access to Spotlight if this is not currently available on your contract.
Admin users have access to Spotlight as a default permission. Add the Spotlight permission to additional roles if you'd like to extend this functionality to other team members.
Roles & Permissions > Role > Access to Reports, Spotlight
Selecting Conversations
To use Spotlight, navigate to the Imported Contacts page and filter your conversation sample to the specific group you want to analyse.
Example: If you're seeing high negative sentiment and low xNPS scores on conversations tagged with "Processing Delays" you can use Spotlight to uncover what customers are actually saying about your claims process.
Once you've applied your filters, click the Spotlight button to analyse these conversations.
Note: Spotlight analyses up to 1,000 conversations. If your filtered set contains more than 1,000 conversations, Spotlight will analyse a representative sample.
Running Spotlight
Once you click on Spotlight a slide out panel will open to indicate the analysis process has begun. Spotlight runs in the background — even if you navigate away from the page. Once the analysis is complete, you'll receive a notification with the results. No need to wait around; carry on with other tasks while Spotlight does its thing.
What's happening:
Spotlight acts as your AI Business Analyst, reviewing the conversation summaries to identify root causes and emerging themes across your customer interactions.
Results are organised into three categories:
Any themes found across conversations will be displayed in 3 categories: Critical Issues, Monitor Closely and Performing Well.
Critical Issues — Surfaces urgent issues that need immediate attention. Think: service breakdowns, at-risk customers, recurring complaints that could lead to churn.
Monitor Closely — Identifies emerging trends before they become problems. Think: minor friction points, process hiccups, early warning signs in customer experience.
Performing Well — Celebrates what's working well so you can do more of it. Think: smooth resolutions, standout service moments, processes worth replicating.
If multiple trends are found these will be listed within the appropriate category
It may be that there are no conversations that fit within a category because there isn't a consistent trend or pattern (for example, if you're looking at negative, detractor conversations it may be there aren't areas of note in the performing well category)
Understanding your results
Multiple themes may appear within each category if different patterns are identified
Some categories may be empty if no consistent patterns exist (for example, if you're analysing negative feedback, "Performing Well" may be empty)
Each theme shows the number and percentage of conversations it represents, helping you prioritise where to focus
Viewing Evidence
Each theme includes real conversation examples so you can verify the AI's findings and gather supporting evidence for action plans or team coaching.
To view examples:
Click on any theme's example dropdown to see sample conversations
Each example shows:
Conversation Reference (clickable link to the full conversation)
Date of the conversation
Agent who handled it
Snippet from the conversation summary showing why it was categorised here
Download CSV
Export your findings to share with stakeholders, track trends over time, or analyse further.
Click the download button to export all conversations and themes in CSV or JSON format.
Your export includes:
Category/Severity: Critical, Monitor Closely, or Performing Well
Theme: The specific issue or pattern identified (for example, "Policy Confusion")
Percentage: What portion of analysed conversations show this theme (for example, 55%)
Reference: Customer contact reference number
Excerpt: Relevant snippet from the conversation summary