Analytics
Track performance, demand, outcomes, customer experience, and operations for your AI agent
Analytics#
Monitor how your AI agent is performing and where customers need more help. Open Analytics in the agent sidebar.
Analytics Navigation#
| Section | What it shows |
|---|---|
| Performance | Interaction volume, coverage, support mix, channel health, and reputation signals |
| Demand | What customers ask for — offerings, goals, and recurring topics |
| Outcomes | Goals completed, conversions, lost-sale signals, and follow-up results |
| Experience | Customer emotion, satisfaction, effort, and frustration patterns |
| Operations | Alerts, AI paused work, human reply rates, and open operational work |
Filters#
Use the filter bar on any Analytics page:
- Date range and chart granularity
- Data scope — All Data, Live Only, or Imported Only
- Interaction source — All Channels, Messages, Email, Comments, or Reviews
- Platform (when relevant)
- Branch / location (when you use multi-location)
Performance#
The main health check for your agent.
KPI cards#
| Metric | Description |
|---|---|
| Total Interactions | Conversations and interactions in the range |
| Unique Contacts | Distinct customers involved |
| Action Needed | Items that still need attention |
| Reply Coverage | How fully customer messages are being answered |
Sections you may see#
- Support Performance — mix of AI and human involvement, feedback where available
- Improvement Proof — signals that setup and training are moving quality in the right direction
- Channel Health — how each connected channel is performing
- Interaction Trends — volume over time
- Reputation Signals — review-related performance when Google Business is connected
- Customer Locations — where customers are located when that data is available
Exact cards can vary based on your channels and data volume.
Demand#
One workspace for what customers want:
- Offerings — products and services customers ask about
- Customer goals — what they were trying to accomplish
- Topics — recurring themes in conversations
Use Manage (or similar controls on the page) to refine how offerings, goals, and topics are defined so reports stay useful for your business.
Outcomes#
Track results, not just traffic:
- Goals completed vs detected
- Conversions and order/booking-style outcomes when detected
- Lost-sale or drop-off reasons
- Follow-up funnel results
Use Outcomes together with follow-up Workflows when you re-engage quiet leads.
Experience#
Understand how conversations felt:
- Sentiment and emotion patterns
- Satisfaction and effort indicators
- Frustration or escalation patterns
Use this page to spot tone issues, training gaps, and channels that need more human coverage.
Operations#
An ops-focused view of work in motion:
- Alerts and items needing review
- Conversations where AI is paused
- Human reply rates
- Open operational work your team still owns
Pair Operations with Activity when you need to act on a specific thread.
Tips#
- Start on Performance each week for coverage and action-needed items.
- Use Demand before adding sources or procedures — train for what people actually ask.
- Check Operations when the team feels busy but isn't sure why.
- Filter to Live Only when you want production channel data without imports or tests.