CX Intelligence helps your team automatically analyze customer conversations for topics, satisfaction scores, and actionable insights. Before your team starts using it, complete the guided setup so CX Intelligence can better understand your business, customers, and the types of insights that are most useful to your team.

Before you begin

Before setting up CX Intelligence, make sure:

• Your workspace has access to CX Intelligence
• You have permission to configure CX Intelligence settings
• Your workspace has customer conversations available for analysis
• You have prepared basic business information, such as your company website, business type, target teams, and analytics goals

Access CX Intelligence settings

To open CX Intelligence settings:

1. Click the icon in the left navigation bar to open CX Intelligence 
2. In the left-side panel, click “Settings”
   

From Settings, you can configure business context, enable or disable CSAT scoring, manage calibration examples, and prepare historical data for analysis.

Step 1: Add your business context

Business context helps CX Intelligence generate more relevant analysis for your company. It tells the AI what your business does, who your customers are, and what outcomes your team cares about.

If this is your first time using CX Intelligence, you may be directed to the Topics page, where the setup flow starts automatically. 

You can also start or rerun setup from Settings by clicking “Update business profile”

You will be asked to enter or confirm the following:

• Company domain: Your company website URL, which helps CX Intelligence learn more about your business
• Business type: The industry that best describes your company, such as e-commerce, SaaS, healthcare, or finance
• Goals: What you want to learn from customer conversations, such as customer pain points, product feedback, or purchase intent
• Target teams: The teams that will use CX Intelligence, such as Customer Support, Sales, Product, or Marketing
• Priority outcomes: The business outcomes you want to improve, such as customer satisfaction, sales, churn, or support volume
• Additional context: Any extra details that may help CX Intelligence understand your business, such as seasonal trends, common customer issues, or industry-specific terms
  

Step 2: Review your business context

After the research is complete, review the AI-generated business context on the onboarding page.

This context is used across CX Intelligence, including Topics, CSAT, and Ask, to make analysis more relevant to your business.

Before confirming, check that:

• Your business summary is accurate
• Your selected goals match what your team wants to learn
• Your target teams reflect the teams that will use the insights
• Your priority outcomes match your current business focus
• Any company-specific terms, seasonal patterns, or common issues are included
• You can update this context later from Settings if your business, products, teams, or priorities change.

Step 3: Review suggested topics

After you confirm your business context, CX Intelligence may suggest topics based on your business information and conversation data.

Topics help classify customer conversations into meaningful categories, such as billing questions, delivery issues, product feedback, purchase intent, or support requests. This makes it easier to understand why customers contact your team and how those patterns change over time.

For each suggested topic, you can:

1. Click the icon next to the topic you want to manage
2. Select one of the following actions:

• View conversations: Review the conversations classified under this topic
• Edit: Update the topic name, description, or subtopics
• Duplicate: Create a copy of the topic and adjust it for a similar use case
• Hide topic: Remove the topic from your active topic view without deleting it
• Delete: Permanently remove the topic from CX Intelligence
  

You can also create your own topics from the Topics page by clicking “Create topic”

Step 4: Enable CSAT scoring

1. Click the icon in the left navigation bar to open CX Intelligence 
2. In the left-side panel, click “Settings”

3. Toggle on “CSAT”

When CSAT scoring is toggled on, CX Intelligence will analyzes new conversation sessions and assign a satisfaction score on a 0–5 scale. You can then review scores, trends, and low-scoring conversations from the CSAT page.

Step 5: Manage CSAT calibration examples

Calibration examples help CX Intelligence understand your company’s scoring expectations.

You can use examples to show what different satisfaction scores should look like for your business. For example, your team may want to score unresolved refund requests differently from resolved delivery issues, even if both conversations contain negative sentiment.

To manage calibration examples:

1. Click the icon in the left navigation bar to open CX Intelligence 
2. In the left-side panel, click “Settings”
3. Toggle on “CSAT”
   
4. Click “Manage examples” 
   
5. Review the available examples
   
6. Turn examples Active or Inactive. Only active examples influence scoring. You can have up to 10 active examples at a time.
   
7. Click theicon to edit the reasoning
   
8. Click the icon to delete examples that are no longer relevant
    

Step 6: Refresh historical data

Use “Refresh data” under Settings to analyze past conversations for selected CX Intelligence features. This helps CX Intelligence apply your latest setup to existing data, especially when you are setting it up for the first time or after updating your business context, topics, CSAT settings, or custom objects.

 Click “Analyze again” to start refreshing the data

Start exploring your insights

Once setup is complete and data starts processing, your team can start using CX Intelligence from different sections:

• Use Topics to understand what customers are talking about
• Use CSAT to monitor satisfaction trends
• Use Ask to ask questions about your conversation data in plain language
  
  
  
  

  Data refreshness

  CX Intelligence does not update every section in real time. Some results appear minutes after a conversation session ends, while other metrics and summaries may take up to 24–36 hours to refresh.
  
   

  For the most accurate results, use CX Intelligence to review recent trends and patterns instead of monitoring live activity. Questions about today’s activity in Ask may return incomplete results until the next data update is complete.

   

Best practices

To get better results from CX Intelligence:

• Keep your business context accurate and up to date
• Review suggested topics before accepting them
• Use clear topic names and descriptions
• Add calibration examples that reflect your team’s CSAT expectations
• Keep only relevant calibration examples active
• Use data backfill after major setup or business context changes
• Review trends over time instead of relying on a single conversation or data point
• Treat AI-generated insights as a starting point for investigation and team review