The Analytics page gives you a data-driven view of how your AI agent and team are performing. Track session volume, customer satisfaction, response times, and more — all with flexible date ranges and CSV export. This page is available to Admins and Owners.
Date Range
Use the dropdown at the top to select your reporting window:
- Last 7 days — quick snapshot of the past week
- Last 14 days (default) — two-week trend view
- Last 30 days — monthly performance
- Last 90 days — quarterly overview
All metrics and charts update immediately when you change the range.
Volume Stats
Four cards at the top show your headline numbers for the selected period:
| Metric | What It Measures |
|---|---|
| Total Sessions | All chat conversations started by visitors |
| Bookings | Meetings booked through the chat widget (via Google Calendar, Calendly, etc.) |
| Resolved | Conversations marked as resolved by an agent or the AI |
| Messages | Total messages exchanged (customer + AI + agent) |
Performance KPIs
Four KPI cards below the volume stats measure quality and efficiency:
CSAT (Customer Satisfaction)
The percentage of rated conversations that received 4 or 5 stars. A CSAT of 85% means 85 out of every 100 rated conversations were positive. Below it, you'll see the total number of ratings collected.
Avg First Response
The average time (in minutes) between a visitor's first message and the first AI or agent reply. Lower is better — this measures how quickly your system responds. The sample size is shown below (up to 500 recent sessions are measured).
Avg Resolution
The average time (in minutes) from when a conversation starts to when it's marked as resolved. This captures the full lifecycle — from first contact to problem solved.
Deflection Rate
The percentage of resolved conversations that were handled entirely by the AI, without any human agent joining. A high deflection rate means your AI is effective — it's resolving issues on its own without needing escalation. This is one of the most important metrics for measuring your AI's ROI.
Session Volume Chart
A daily area chart shows session volume over the selected date range. Each point represents one day. Use this to:
- Spot traffic patterns (weekday vs weekend, time of year)
- Identify spikes (after a product launch, marketing campaign, or outage)
- Track growth trends over time
Hover over any point to see the exact date and session count.
CSV Export
Click Export CSV to download a spreadsheet with all metrics for the selected date range. The CSV includes:
- Summary section — all KPIs in one row (total sessions, resolved, messages, bookings, CSAT, avg response time, avg resolution time, deflection rate)
- Daily breakdown — one row per day with sessions, resolved, messages, and bookings
Import this into Google Sheets, Excel, or your BI tool for custom charts, team reports, or historical tracking.
How Metrics Are Calculated
| Metric | Calculation |
|---|---|
| CSAT | (Ratings of 4 or 5) / (Total ratings) x 100 |
| Avg First Response | Average of (first AI/agent message time - first customer message time) across sampled sessions |
| Avg Resolution | Average of (resolution time - session creation time) across resolved sessions |
| Deflection Rate | (Resolved sessions without agent) / (Total resolved sessions) x 100 |
Tips
- Check analytics weekly — a 14-day view gives you enough data to spot trends without being overwhelmed by noise.
- Track deflection rate over time — as you add more documents and improve your knowledge base, this should trend upward. If it drops, check Knowledge Gaps for new unanswered questions.
- Use CSAT to measure quality — a high deflection rate is only good if CSAT stays strong. If CSAT drops while deflection rises, the AI may be resolving conversations too aggressively without fully helping customers.
- Export for team reviews — download the CSV for weekly team meetings. Combine it with Reviews feedback for a complete picture of support quality.
- Watch first response time — if it creeps up, your AI may be slow (check document processing) or your team may need to join conversations faster.
- Use 90-day view for trends — quarterly data smooths out daily fluctuations and shows the real trajectory of your support operation.