AI outbound calls are not a replacement for human coordinator calls — they are a volume tool that handles a different segment of your lead pipeline at a different cost. Using AI for calls that require human judgment wastes an opportunity that a coordinator could have handled better. Using a coordinator for calls that AI can handle efficiently at $0.30 each misallocates clinical expertise that costs $25–35 per call to deploy. The decision framework below tells you which calls belong where.
The Decision Matrix
| Scenario | AI Outbound | Human Outbound |
|---|---|---|
| Cold lead, 90+ days since contact, no response to 3 prior attempts | ✓ | |
| No-show appointment, same-day recovery | ✓ (Touch 1–2) | ✓ (Touch 3) |
| Pre-screened lead, qualified, slow to schedule | ✓ (initial re-contact) | ✓ (if AI unanswered) |
| Borderline eligibility patient needing clinical judgment | ✓ | |
| Patient with stated safety concern or symptom | ✓ (immediate) | |
| Physician referral follow-up | ✓ | |
| Enrolled participant logistics (visit reminders) | ✓ | |
| Newly screened patient, enrollment decision conversation | ✓ | |
| Study completion follow-up and review request | ✓ (initial) | ✓ (if no response) |
The Cost Comparison
- AI outbound call: $0.10–0.50 per call including platform cost. Can run 200 calls simultaneously. No capacity limit. No emotional fatigue after call 15.
- Human coordinator outbound call: $12–22 per call when accounting for salary, benefits, and time (including preparation and logging). Maximum 20–30 effective outbound calls per coordinator per day. Quality decreases with repetitive cold calling.
The 40:1 cost ratio makes AI outbound clearly superior for volume re-engagement. The human judgment requirement makes human outbound irreplaceable for clinical, relational, and complex conversations.
How to Split Your Lead List
Each month, segment your outbound call list into two buckets:
- AI bucket: Cold leads (60+ days), no-show day-1 recovery, visit reminders, review requests. Run these through AI outbound first. Any who express interest (press 1, reply, call back) move to the human bucket.
- Human bucket: Warm leads (interested but not yet scheduled), borderline eligibility cases, physician referrals, current participants with questions. All get coordinator calls with appropriate context from the CRM.
Measuring the Split Performance
Track monthly: AI outbound conversion rate (AI calls → human-bucket transfer), human outbound conversion rate (human calls → scheduled visit), and overall cost-per-scheduled-visit for each channel. Adjust the split based on data. If AI is converting cold leads at 15% to human-bucket transfers, it is delivering high value. If it is converting at 3%, the script or targeting needs adjustment before the next campaign.
48-Hour Action List
- Hour 1: Categorize your current outbound call list using the decision matrix above. Count how many calls belong in the AI bucket vs. the human bucket. Calculate how much coordinator time would be saved by shifting the AI bucket to an AI outbound system.
- Hour 2: Calculate your current human outbound cost-per-scheduled-visit for cold leads: coordinator hours spent on cold calls this month × hourly cost ÷ scheduled visits generated. This is the benchmark AI outbound will compete against.
- Hour 3: Run a 20-call AI outbound pilot on your cold lead bucket using a platform like Bland AI or VAPI. Track cost, answer rate, and conversion to human-bucket transfer.
- Day 2: Compare the AI pilot results to your human cold-call benchmark. If AI produces scheduled visits at lower cost-per-visit — which it typically does for leads 60+ days cold — expand the AI bucket. If human outperforms, examine why: possibly your cold list is warmer than expected or the AI script needs improvement.