Skip to content

Where Insights Appear

Conversation Analytics results show up in several places so different roles can take action quickly.

1) Conversation details

For an individual conversation (call/chat/email/ticket), users can view:

  • the transcript/thread
  • extracted insight values (scores, categories, entities, summaries)
  • explanations that justify the extracted values

This view is ideal for:

  • coaching and QA reviews
  • investigating a specific customer issue
  • validating whether an insight definition matches your business expectations

Tip: Explanations are designed to help reviewers understand the why, not just the final score/category.

Call Details showing AI-generated call summary and insights

Figure: Call Details view with AI-generated summary. The Analytics tab displays extracted values such as CSAT score and sentiment along with explanations.

2) Dashboards

Dashboards are built from stored insight values (Custom Fields) and typically include:

  • trend lines over time
  • averages (for numeric metrics like CSAT)
  • distributions (e.g., % dissatisfied / neutral / satisfied)
  • breakdowns by team/agent/queue (where enabled)

A key workflow is drilldown:

  • click a bucket (e.g., “Dissatisfied”) to navigate to the conversations behind it
  • review a sample of those conversations with explanations
  • take action (coaching, process changes, knowledge base updates)

CSAT Dashboard with distribution buckets

Figure: CSAT Dashboard showing score distribution, trend over time, and per-user breakdown. Click any bucket to drill down to matching conversations.

3) Search and filters

Users can search and filter conversations using insight values, for example:

  • CSAT < 3
  • Churn risk = High
  • Competitor mentioned = Yes
  • Topic contains “Billing”
  • Sentiment = Negative

Because insights are stored as structured fields, you can filter precisely (ranges, categories, dates), not just keyword search.

Call list with search filters applied

Figure: Search results filtered by insight values. Users can combine filters (date range, CSAT score, sentiment, etc.) to find specific conversations.

4) Reports and exports (when enabled)

Some deployments enable reporting or exporting insights for:

  • operational reporting
  • downstream BI
  • integrations and automation

Reports generally work best when insights are stored in well-defined fields with consistent types and allowed values.

Next: How to interpret and validate results

If you are evaluating or rolling out insights, you’ll want to understand quality expectations and best practices: