Analytics PRO

The Analytics module transforms raw conversation data into actionable insights. Track conversation funnels, measure customer satisfaction, detect anomalies, build custom dashboards, and schedule automated reports — all from a single interface.

Summary Cards

The analytics dashboard opens with four summary cards that provide an at-a-glance overview of your instance's activity:

Total Messages — The total number of messages sent and received across all channels and agents in the selected time range
Conversations — The number of distinct conversation sessions initiated during the period
Avg Response Time — The mean time between a user message and the agent's reply, measured in seconds
Peak Hour — The hour of day with the highest message volume, helping you understand when your users are most active

Each card shows the current value alongside a percentage change compared to the previous period, so you can instantly see whether metrics are trending up or down.

Charts

Four built-in charts visualize your conversation data across different dimensions:

Messages Over Time — A line chart showing message volume by day, revealing trends and seasonal patterns
Requests by Hour — A bar chart breaking down activity by hour of day, useful for capacity planning and business hours decisions
Messages by Agent — A stacked chart showing how message volume distributes across your deployed agents
Daily Conversations — A bar chart of unique conversation sessions per day, distinct from raw message count

All charts support 7-day, 14-day, and 30-day time ranges. Select the range from the dropdown at the top of the page, and every chart and summary card updates to reflect the selected window.

Conversation Funnels

Funnels track how conversations progress through five stages:

Started — User initiated a conversation
Engaged — User sent at least two messages
Substantive — Conversation reached meaningful depth (5+ message exchanges)
Resolved — The conversation reached a resolution (user confirmed or conversation ended naturally)
Satisfied — User provided positive feedback or a high CSAT rating

The funnel visualization shows the drop-off rate between each stage. If a large percentage of users start conversations but few reach the "Engaged" stage, your agents may need better opening responses. If most conversations engage but few resolve, the agents may need more domain knowledge via the Knowledge Base.

Tip: Funnel data pairs well with conversation path analysis. Check the paths view to understand where users drop off, then use funnels to measure whether your improvements change the drop-off rates.

CSAT Dashboard

The Customer Satisfaction (CSAT) dashboard collects and visualizes satisfaction ratings from your users. When CSAT collection is enabled, users are prompted to rate their experience at the end of a conversation.

The dashboard displays:

Overall CSAT score as a percentage
CSAT trend over time (daily scores plotted on a line chart)
Score distribution breakdown (1-5 star histogram)
Per-agent CSAT scores so you can identify which agents perform best
Per-channel CSAT scores to compare satisfaction across communication channels

Conversation Paths Visualization

The paths view renders a Sankey-style diagram showing how conversations flow through different topics, intents, and outcomes. Each node represents a conversation state, and the width of each link represents the volume of conversations that followed that path.

Use this to identify the most common conversation flows, spot unexpected branching points, and discover paths that lead to poor outcomes. Conversation paths are generated automatically from your message data — no manual tagging required.

Anomaly Detection

The anomaly detection engine monitors your key metrics continuously and alerts you when values deviate significantly from their expected range. Anomalies are detected using statistical baselines computed from your historical data.

When an anomaly is detected, it appears as a highlighted point on the relevant chart along with a notification in your dashboard. Examples of detected anomalies include:

A sudden spike in error rates
An unusual drop in conversation volume
Response times exceeding historical averages by more than two standard deviations
CSAT scores dropping below the rolling average

Anomaly alerts can be forwarded to external services via Webhooks.

Custom Dashboards

Build your own dashboard layouts by arranging charts in a drag-and-drop grid. Add, remove, resize, and reposition charts to create a view tailored to your needs. Custom dashboards are saved to your account and persist across sessions.

You can create multiple custom dashboards for different purposes — one for daily operational monitoring, another for weekly executive reporting, and a third focused on a specific agent or channel.

Cohort Analysis

Cohort analysis groups users by when they first interacted with your agents and tracks their behavior over subsequent periods. This reveals retention patterns and helps you understand whether changes to your agents improve long-term engagement.

Cohorts are defined by week or month of first interaction. The retention table shows what percentage of each cohort returned in subsequent periods.

Scheduled Reports

Automate your reporting with scheduled email delivery. Configure a weekly report that includes your summary cards, top-level funnel metrics, and CSAT scores. Reports are delivered as formatted HTML emails with embedded charts.

Set the delivery day and time, add multiple email recipients, and choose which metrics to include. Scheduled reports use the same data source as the live dashboard, ensuring consistency between what you see on screen and what arrives in your inbox.

Channel Analytics

Channel Analytics provides per-channel breakdowns of message volume, response times, and engagement metrics. Compare performance across your connected channels to understand where your users prefer to interact and which channels deliver the best experience.

Auto-Responses

Configure automatic replies that trigger on keyword matches or during specific time windows (such as outside business hours). Auto-responses reduce response latency for common queries and ensure users always receive an acknowledgment, even when your agents are processing complex requests.

Export Data

Export your analytics data for use in external tools. All charts and tables support export, and the raw data underlying each visualization is available for download. Use exported data for custom analysis, compliance reporting, or integration with business intelligence platforms.

Want real-time logs instead? Analytics provides aggregate insights over time. For real-time operational visibility, use the Logs Explorer.