Skip to main content
The BotSmith analytics dashboard gives you a real-time snapshot of how your agents are performing. Use it to understand usage trends, spot top-performing agents, and identify opportunities to improve responses.
Analytics data refreshes approximately every 60 seconds.

​
Dashboard Overview

Navigate to Analytics in the sidebar to reach the main analytics page. At the top you’ll find five key metric cards, each updated in real time: Use the 7 Days / 30 Days / 90 Days selector in the top-right corner of the page to adjust the time window for all trend data shown below the metric cards.

​
Per-Agent Analytics

To drill into a single agent’s performance, you have two options:
  • From the dashboard — Click the agent’s name or row in the agent list.
  • From the agent builder — Open the agent, then select the Analytics tab.
The per-agent view includes:
  • Messages over time — A line chart showing how many messages the agent exchanged each day within your selected date range.
  • Conversations per day — A daily breakdown of new chat sessions.
  • Most common user queries — A bar chart ranking the questions your visitors ask most frequently, helping you spot gaps in your agent’s knowledge base.

​
Trend Charts

The Trends section aggregates data across all your agents so you can see the bigger picture. It plots:
  • Message volume — Total messages sent and received each day.
  • Conversation count — New sessions started each day.
  • Average response time — The mean time (in seconds or milliseconds) between a visitor’s message and your agent’s reply, calculated from actual message timestamps.
Use the date range selector to view the last 7, 30, or 90 days. The charts update automatically when you change the window.

​
Advanced Analytics

Advanced analytics are available on the Starter plan and higher. Upgrade your plan to unlock these charts inside each agent’s Analytics tab.
The advanced analytics panel for each individual agent provides a deeper set of metrics:
  • Message volume trends — Daily breakdown of conversations and messages, with per-integration channel detail (Dashboard, Widget, Telegram, Slack, Discord, WhatsApp, Instagram, Messenger, MS Teams).
  • Top asked questions — A ranked list of the most frequently submitted user questions, normalized and counted across all conversations.
  • Satisfaction distribution — A pie chart of 1–5 star ratings submitted by visitors after conversations, including average rating and total rating count.
  • Performance metrics — Fastest, average, and slowest agent response times in seconds, measured from actual user-to-assistant message pairs.
  • Response time trend — Average response time plotted day by day over your selected period.
  • Hourly activity distribution — A bar chart showing message volume by hour of the day (00:00–23:00), so you can identify your agent’s busiest times.

​
Interpreting Metrics

Here is what each top-level number means in practice:
  • Total Conversations — Every time a visitor opens a new chat session with one of your agents, that counts as one conversation. Returning visitors who start a new session each time also each count separately.
  • Total Messages — Every individual exchange counts as one message, whether it is a visitor’s question or your agent’s reply. A conversation with five back-and-forth turns contributes ten messages to the total.
  • Active Agents / Total Chatbots — The dashboard card shows both numbers together (for example, 3 / 5). Only agents explicitly set to Active status are counted in the active figure. Agents in Draft or Inactive status are excluded from the active count, even if they have historical conversation data.

​
API Access

You can fetch dashboard analytics programmatically. See the Analytics API Reference for full schema details.
Replace YOUR_TOKEN with your API token from Account → API Keys. To fetch trend data across all your agents, use the trends endpoint with an optional days query parameter (1–365, default 30):