how-to
Updated What metrics should I track after deploying an AI SDR?
The AI SDR metrics dashboard
| Category | Metric | Target |
|---|---|---|
| Conversation quality | Retrieval-miss rate | <5% |
| Conversation quality | Containment rate (resolved without human) | >70% |
| Conversation quality | Post-conversation CSAT | >4.2/5 |
| Pipeline | Meeting-book rate on qualified traffic | 6–12% |
| Pipeline | Show rate on booked meetings | >80% |
| Pipeline | Opportunity-conversion from booked | 30–50% |
| Speed | First-response time | <3 seconds |
| Speed | Time to booked meeting | <60 seconds |
| Economics | Cost per meeting (fully loaded) | $8–$25 |
| Economics | Pipeline-to-cost ratio | >6:1 within 6 months |
Report weekly for the first 90 days, monthly thereafter. See the ROI calculator and ROI explainer for the full model.
Related questions
What's a bad meeting-book rate for AI SDR?
Under 4% on qualified traffic. That signals either poor conversation quality, wrong routing rules, or a knowledge base that doesn't cover the questions prospects are asking.
How do I attribute pipeline to the AI SDR?
Tag every conversation with a source ID that flows into the CRM opportunity record. Report on closed-won revenue where the first-touch is the AI SDR conversation. Without source tagging, the ROI case is unprovable.