Skip to main content

Insights

The Insights page provides comprehensive analytics and visualizations to help you understand how users interact with your branchly application. Through a collection of interactive charts and metrics, you can track engagement patterns, analyze conversation topics, monitor sentiment, and gain geographic insights into your user base.

Overview​

Insights offers real-time analytics across multiple dimensions of user behavior. All charts and metrics are interactive and can be filtered by time period to focus on specific date ranges that matter most to your business.

Time Period Selection​

Use the date picker at the top of the page to customize the time frame for all charts. Available preset options include:

  • Last 30 Days: Rolling 30-day window from today
  • This Month: Current calendar month to date
  • Last Month: Previous complete calendar month
  • Last 6 Months: Rolling 6-month window
  • Last 12 Months: Rolling 12-month window
  • This Year: Current calendar year to date
  • Custom Range: Select specific start and end dates

Note: For long time periods (greater than 6 months), session data is aggregated into weekly buckets instead of daily buckets for better performance and visualization clarity.

Available Charts​

Active Sessions by Interaction Type​

This time-series chart visualizes the volume of active sessions over time, broken down by interaction type:

  • Chat: Sessions involving conversational AI interactions
  • Search: Sessions using semantic search functionality
  • Navigation: Sessions where users navigated through the knowledge base structure
  • Form Submissions: Sessions where users submitted forms

Data Aggregation: Daily buckets for periods ≤6 months, weekly buckets for longer periods.

Answer Types​

This chart categorizes AI-generated responses by type, providing insights into how the AI handles user queries:

  • Complete Answers: Questions fully answered by the AI
  • Follow-Up Question: Cases where the AI asked for clarifying information
  • Small Talk: Basic small talk with the user, i.e. replying to greetings
  • No Knowledge: Queries the AI couldn't address with available information and mentions that it does not know the answer.
  • Out of Scope: Questions outside the defined scope

Conversation Topics (Tags)​

Automatically generated tags categorize conversations by topic, helping you understand what users are asking about. The AI analyzes each conversation and assigns relevant topic tags.

Use Cases:

  • Discover trending topics and common user concerns
  • Identify content opportunities for your knowledge base
  • Understand seasonal or event-driven inquiry patterns
  • Prioritize documentation improvements

Display: Shows the most frequent tags with their occurrence counts.

Clicked URLs​

Tracks which links users click within your branchly interfaces, including:

  • Links in search results
  • URLs in AI responses
  • Navigation node links
  • External resource links

Use Cases:

  • Identify most valuable content and resources
  • Validate that users find what they're looking for
  • Discover popular content paths
  • Optimize content placement and recommendations

Interaction Sources​

Shows where users are engaging with your branchly interfaces across your website, i.e. page URLs where interfaces are embedded.

Use Cases:

  • Understand which pages drive the most engagement
  • Optimize interface placement on high-traffic pages
  • Identify underutilized embedding opportunities
  • Track campaign effectiveness if interfaces are on landing pages

Top Search Queries​

Displays the most frequent search terms and questions users enter, providing direct insight into user intent.

Use Cases:

  • Identify common user needs and pain points
  • Optimize knowledge base content for popular queries
  • Discover gaps in your content coverage
  • Improve SEO and content strategy
  • Inform product development based on user needs

Sentiment Analysis​

Analyzes the emotional tone of chat interactions, categorizing them as:

  • Positive: Satisfied, happy, or successful interactions
  • Neutral: Informational or transactional exchanges
  • Negative: Frustrated, confused, or unsuccessful interactions

Use Cases:

  • Monitor overall user satisfaction
  • Identify problematic interaction patterns
  • Correlate sentiment with specific topics or time periods

Locales​

Shows the distribution of user language and region preferences based on your website lang tags.

Format: Displays as language-country codes (e.g., en_US, de_DE, fr_FR)

Identified Languages​

Detects and categorizes the actual languages used in chat interactions, regardless of the locale settings. This provides insight into the real language distribution of your chat users.

Devices​

Breaks down user sessions by device type:

  • Desktop: Traditional computer browsers
  • Mobile: Smartphones
  • Tablet: Tablet devices

Geography​

A chart visualization showing where your users are located based on IP address data.

Privacy: Geographic data is aggregated and anonymized, using approximate IP ranges to protect individual user privacy while still providing valuable regional insights.