By Steve Baylis
Print this. Put it next to your monitor. Look at it side-by-side with your live scorecard for the first two weeks.
The 6 metrics that actually matter
1. AI response time
What it measures: How fast AI replied to inbound leads (median, last 7 days).
Good: Under 12 seconds. Bad: Over 30 seconds — usually means a phone-system or webhook problem. What to do if it’s bad: Open Settings → Phone System and check that all numbers are registered.
2. AI conversation continuation rate
What it measures: % of inbound leads that engaged in 2+ messages with AI.
Good: 60%+ for fresh leads. Bad: Under 40% — usually means key notes are wrong or hours are misconfigured. What to do if it’s bad: Re-read your key notes (Library → AI Agent Key Notes Template). Adjust tone.
3. Appointment book rate
What it measures: % of AI conversations that ended in a booked appointment.
Good: 25%+ of qualified leads. Bad: Under 15% — escalation rules might be too tight, or AI is qualifying too aggressively. What to do if it’s bad: Review last 20 conversations that didn’t book. Look for pattern — is AI cutting off too early?
4. Show rate (per rep)
What it measures: % of booked appointments that actually showed up.
Good: 70%+ Bad: Under 50% — usually a rep-side problem (no confirm, weak appointment slot). What to do if it’s bad: Coach the rep on confirmation patterns. AI auto-confirms day-before; the rep needs to confirm same-day.
5. Won deals from AI-warmed leads (per rep)
What it measures: Closed deals where AI was the first touch (last 30 days).
Good: Top performers should be at 8+ per month. Bad: Under 3 — coaching opportunity. The rep is getting handoffs but not closing. What to do if it’s bad: Listen to recent conversations on those leads. Often the rep is going too cold too fast on a primed buyer.
6. Stale pipeline (per rep)
What it measures: Leads in the rep’s pipeline with no activity in 7+ days.
Good: Under 10 stale leads. Bad: Over 25 — they’re sitting on a pile of forgotten deals. What to do if it’s bad: Have the rep work through the stale list one shift. Sales drop fast when reps stop touching their pipeline.
The operating principle
The two lowest reps each week need a 1:1.
Look at the scorecard each Monday. Sort by won deals descending. The bottom two get a 30-minute coaching session this week.
Don’t punish — coach. Pull up specific conversations, point at specific moments, ask “what would you do differently?”
That’s the system. Six metrics. Two coaching conversations. Every Monday.
Quick-glance summary
| Metric | Target | Coaching trigger |
|---|---|---|
| AI response time | <12s | >30s |
| Continuation rate | 60%+ | <40% |
| Appointment book rate | 25%+ | <15% |
| Show rate (per rep) | 70%+ | <50% |
| Won deals from AI (per rep) | 8+/mo | <3/mo |
| Stale pipeline (per rep) | <10 | >25 |
AI Dealership Academy · learn.diablo-ai.com