A dental practice was missing roughly one-third of incoming calls. Voicemails piled up overnight. New patients called and got no answer. After-hours calls went nowhere. The front desk was too busy with in-office patients to pick up every ring.
After deploying an AI receptionist, the practice captured 98% of all incoming calls — including 122 after-hours calls in a single month that previously went to voicemail. In 30 days: 417 calls handled, 32 appointments booked by AI, $38,400 in recovered revenue. Average response time: 25 seconds.
Over 90 days across 2 locations: 1,700+ calls, 180+ appointments, 12 new patients, $247,500 in production revenue.
This is the full breakdown — what the practice looked like before, what changed, and the specific numbers from the deployment.
Before: What Missed Calls Looked Like
The practice had one full-time receptionist and one part-time. During office hours, the front desk handled check-ins, check-outs, payments, insurance questions, and incoming calls — all simultaneously.
The result:
During peak hours (Monday mornings, lunch, end of day): 2-3 calls coming in at once. One answered, two to voicemail.
During lunch: Phones forwarded to voicemail. Staff returned to 8-10 messages.
After 5 PM: All calls to voicemail. No coverage evenings, weekends, or holidays.
New patient calls: If they hit voicemail, most did not call back. They called the next practice on their search results.
The practice estimated they were missing 30-35% of all incoming calls. At roughly 400 calls per month, that was 120-140 unanswered calls — many from potential new patients worth $1,000-$1,500 each in first-year revenue.
The Decision
The practice evaluated three options:
| Option | Monthly Cost | Coverage | Books Appointments? |
|---|---|---|---|
| Hire a second full-time receptionist | $3,500 - $5,800 | 40 hrs/week | Yes |
| Human answering service | $800 - $1,500 | Varies | No (messages only) |
| AI receptionist | $800 | 24/7 | Yes (real-time PMS booking) |
The human answering service was ruled out because it only took messages — the patient would still need a callback the next day to actually book. That defeated the purpose for after-hours calls.
Hiring a second receptionist solved the overflow problem during business hours but not after-hours coverage — and cost 4-7x more.
The practice chose AI at $800/month with a 60-day pilot.
Setup
The AI connected to the practice's PMS (Open Dental) and phone system in 48 hours:
Day 1: Integration with Open Dental. Configured appointment types (cleaning, exam, emergency, consult with correct durations), provider schedules, operatory assignments, office hours, and emergency escalation protocol.
Day 2: FAQ configuration (hours, directions, accepted insurance, first visit instructions). Test calls — booked appointments, rescheduled, asked questions, simulated emergencies. Verified everything appeared correctly in Open Dental.
Day 3: Live. After-hours calls answered by AI immediately. Business-hours overflow added in week 2.
Phone number stayed the same. No hardware installed. Patients noticed nothing different except that someone always picked up.
30-Day Results
| Metric | Before AI | After AI (30 days) |
|---|---|---|
| Total monthly calls | ~400 estimated | 417 (all captured) |
| Missed call rate | 30-35% | Under 2% |
| After-hours calls captured | 0 (all to voicemail) | 122 (29% of total) |
| Appointments booked by AI | 0 | 32 |
| New patients from AI bookings | — | 8 |
| Revenue from AI bookings | $0 | $38,400 |
| Average response time | Variable (voicemail or 3+ min hold) | 25 seconds |
| Morning voicemail backlog | 8-10 per day | 0 |
Key observations
122 after-hours calls — This was the most revealing number. The practice had no idea they were receiving this many calls outside business hours. All 122 were captured and either resulted in a booked appointment, a answered question, or a captured lead for morning follow-up.
32 appointments booked directly by AI — These were appointments that the patient booked during the call, without staff involvement. The AI checked Open Dental availability, confirmed the slot, and sent a confirmation text.
$38,400 in recovered revenue — Based on an average appointment value of $1,200. These were appointments that would not have happened without AI answering the calls.
Morning voicemail backlog eliminated — Staff previously spent the first 30-60 minutes of every morning returning overnight calls. With AI handling after-hours, they started the day with a clean slate and could focus on the patients walking in.
90-Day Results (Expanded to 2 Locations)
After the 30-day pilot succeeded, the practice expanded AI to their second location:
| Metric | 90-day result (2 locations) |
|---|---|
| Total calls handled | 1,700+ |
| Appointments booked by AI | 180+ |
| New patients acquired | 12 |
| Production revenue | $247,500 |
The per-location results were consistent — confirming that the 30-day data was not an anomaly. After-hours calls represented 25-30% of total volume at both locations.
What Changed Operationally
Front desk workload
The receptionist went from being buried in phone calls to focusing on in-office patients. Check-ins became less rushed. Patient interactions at the desk improved because the receptionist was not constantly interrupted by ringing phones.
Staff morale
Front desk burnout decreased measurably. The receptionist described the change as "going from drowning to breathing." The constant multi-tasking of phones + patients + payments + insurance became phones handled by AI + patients + payments + insurance — removing the most stressful variable.
Patient experience
Patients calling after hours got immediate help instead of voicemail. New patients booking at 8 PM got a confirmation text within seconds. Existing patients rescheduling on weekends had it done in under a minute. The practice felt responsive and professional around the clock.
No-shows
With AI sending confirmation texts after every booking and two-way reminder texts before appointments, the practice saw a noticeable reduction in no-shows — patients who might have forgotten their appointment got a text reminder with a one-tap confirm or reschedule option.
ROI Breakdown
| Item | Amount |
|---|---|
| Monthly AI cost | $800 |
| Monthly revenue recovered (30-day) | $38,400 |
| Return on investment | 48x |
| 90-day revenue (2 locations) | $247,500 |
| 90-day AI cost (2 locations) | ~$4,800 |
| 90-day ROI | 51x |
Even at half the results — 16 appointments per month instead of 32 — the ROI is 24x. The math works at any reasonable call volume.
What the Practice Would Do Differently
Start with full coverage from day 1. They started after-hours only and added business-hours overflow in week 2. In hindsight, going full coverage immediately would have captured more overflow calls from the first day.
Track new patient source more carefully. Some of the 32 AI-booked appointments were new patients, some were existing. Better source tracking from the start would have made the ROI calculation even clearer.
Expand to the second location sooner. The 30-day pilot proved the concept. Waiting an additional month before deploying at location 2 meant lost call captures at that site.
Key Takeaways
The practice was missing 30-35% of calls without knowing the true volume — especially after hours
After-hours calls were 29% of total volume — a revenue stream that was completely invisible before AI
AI booked 32 appointments in 30 days that would have been lost to voicemail
Setup took 48 hours with no hardware and no phone number changes
The cost ($800/month) was a fraction of the recovered revenue ($38,400/month)
Results scaled consistently when expanded to a second location
FAQs
What PMS was the practice using?
Open Dental. The AI connected directly and booked appointments in real time.
Did the practice keep their receptionist?
Yes. The AI handled overflow and after-hours. The receptionist focused on in-office patients, check-ins, and complex calls.
How were the 32 appointments attributed to AI?
These were appointments booked during AI-handled calls — tracked by the AI system with call recordings and PMS booking timestamps.
What happened to calls AI could not handle?
Warm transfer to staff during business hours with a full context summary. After hours, detailed summary captured for morning follow-up.
Can any practice expect these results?
Results vary by call volume and how many calls are currently being missed. Practices missing 20-30% of calls will see the most immediate impact. The 25-30% after-hours pattern is consistent across practices.
