Run your workspace
Analytics & gaps
The Analytics page (/app/analytics) is your one-stop view
on whether the agent is actually working. Conversation volume,
deflection rate, lead conversion, and the all-important
knowledge gap report.
Top-level metrics
| Metric | What it means |
|---|---|
| Conversations | Distinct chat threads in the period. Counts each visitor's first init, not every message. |
| Messages | Total visitor messages. Average per conversation tells you engagement depth. |
| Deflection rate | Conversations the AI handled end-to-end (no human takeover, no low-confidence flag) divided by total. Higher is better. |
| Leads captured | Visitors who completed the lead form. Conversion shown as % of conversations. |
| Average response time | Median time from visitor message to first token streamed back. Should sit well under the 1s p95 hot-path budget. |
Widget funnel
The funnel card answers the question the top-level metrics can't: is anyone even seeing the widget? Three steps, each counted once per page load over the selected window:
| Step | Counted when |
|---|---|
| Shown | The widget mounted on a page view (the widget.ready beacon). Impressions have been recorded since the beacon shipped, so history predates the funnel card itself. |
| Opened | The panel actually opened on that page load — a click on the launcher or bubble, an auto-open, or a behavior trigger all count once. |
| Conversations | The same conversation count as above — threads with real activity, excluding playground and health-check traffic. |
The percentages between steps are where to look: a low Opened / Shown ratio points at the teaser or auto-open settings; a low Conversations / Opened ratio points at the greeting and starter questions. A site with modest traffic can have a perfectly healthy funnel and still produce few conversations in absolute numbers — the funnel makes that distinction visible.
The gap report
Whenever a turn comes back with low_confidence=true, a
DetectGapJob is dispatched. The job clusters similar gaps
using semantic similarity on the visitor's question, then opens a row in
the Gaps table. Each row shows:
- The cluster's representative question.
- How many times it has been asked.
- Sample conversation links so you can read the full context.
- An Add source button that pre-fills the source modal with a search query.
Closing the loop: read the gap → add a knowledge source that answers it → next visitor with that question gets a confident answer.
Per-agent vs. workspace
The page defaults to all agents in the workspace; the agent picker at the top filters to one. Useful when you have one agent per surface (marketing, help center) and want to compare deflection rates.
Time ranges
Pick a window: 24h, 7d, 30d, or a custom range. Numbers update without a full reload — Inertia partial visit. For longer trends, the chart at the top is a 30-day rolling view.
Export
The CSV export pulls everything in the current filter — conversations, messages, leads, gaps. One row per record. Columns are stable so you can automate downstream reporting.
Citation effectiveness
A small table shows which knowledge sources were cited most often and which never get cited. Sources with zero citations in 30 days are candidates for deletion or reindexing.
Experiment results
If you have an active A/B experiment on behavior rules (see Behavior rules & triggers), its conversion delta appears as a callout at the top of the page until you call it (declare a winner). The page-level chart respects the experiment split too — you can see whether the variant lifted overall deflection.