Skip to main content

Chat

The Signal Chat feature uses AI to help you analyze sessions, answer questions, and get insights from your session data.

Getting Started

Accessing Chat

  1. Open the Signal app
  2. Click Chat in the sidebar
  3. A new chat conversation will start automatically

Chat Interface

The chat interface includes:
  • Message input - Type your questions at the bottom
  • Conversation history - View previous messages and responses
  • Session context - Chat can reference specific sessions
  • AI responses - Get insights and analysis from AI

Using Chat

Basic Questions

Ask questions about your sessions:
How many sessions did we have yesterday?
What are the most common errors?
Show me sessions with payment failures

Session Analysis

Ask about specific sessions:
Analyze session abc123
What happened in session xyz789?
Why did the user leave in session def456?

Trend Analysis

Ask about trends and patterns:
How has user engagement changed this week?
What are the top user actions?
Show me conversion trends

Filtering Sessions

Ask to filter sessions:
Show me sessions from angry users
Find sessions with checkout errors
List sessions from last week with bugs

Chat Features

Session Context

When viewing a session replay:
  1. Open a session from the Replays page
  2. Click Open in Chat or navigate to Chat
  3. Chat automatically includes session context
  4. Ask questions about that specific session

Conversation History

  • Previous chats - Access past conversations
  • Continue conversations - Pick up where you left off
  • New chat - Start fresh conversations

AI Insights

Chat provides:
  • Session summaries - Quick overviews of sessions
  • Error analysis - Identify and explain errors
  • User behavior - Understand user actions
  • Recommendations - Suggestions for improvements

Example Queries

Session Analysis

Q: What happened in session abc123?
A: [AI analyzes the session and provides a summary]

Q: Why did the user leave?
A: [AI identifies the reason based on session data]

Q: Were there any errors?
A: [AI lists all errors encountered]

Trend Questions

Q: How many sessions did we have this week?
A: [Shows session count and trend]

Q: What's our average session duration?
A: [Provides metric and comparison]

Q: What are the top 5 events?
A: [Lists most frequent events]

Filtering

Q: Show me sessions with payment failures
A: [Lists matching sessions with links]

Q: Find angry users from last week
A: [Filters sessions by auto tags]

Q: Sessions with checkout errors
A: [Shows relevant sessions]

Best Practices

Asking Effective Questions

  • Be specific - “Sessions with payment errors” vs “errors”
  • Include context - Mention time ranges or filters
  • Use natural language - Chat understands conversational queries
  • Reference sessions - Include session IDs when asking about specific sessions

Using Session Context

  • Open sessions first - View session before asking questions
  • Ask follow-ups - Build on previous responses
  • Reference details - Mention specific events or actions

Getting Insights

  • Ask “why” - Understand root causes
  • Request comparisons - Compare time periods or segments
  • Get recommendations - Ask for improvement suggestions

Troubleshooting

Chat Not Responding

  • Check connection - Ensure you’re connected to the internet
  • Wait for response - AI analysis may take a few seconds
  • Try rephrasing - Restate your question differently

Incorrect Answers

  • Provide more context - Include session IDs or time ranges
  • Be more specific - Narrow down your question
  • Check data - Verify sessions exist and have data

Session Not Found

  • Verify session ID - Check the session ID is correct
  • Check permissions - Ensure you have access to the project
  • Confirm existence - Verify the session exists

Tips

Quick Analysis

Use chat for quick insights without navigating multiple pages:
"Summarize yesterday's sessions"
"Find all checkout errors"
"What's our conversion rate?"

Deep Dives

Ask follow-up questions to dive deeper:
Q: Show me payment failures
A: [Lists sessions]
Q: What's the common cause?
A: [Analyzes patterns]
Q: Which users are affected?
A: [Provides user breakdown]

Comparisons

Compare different time periods or segments:
"Compare this week to last week"
"Show me differences between mobile and desktop"
"What changed after the update?"