Tellet.net

Case Study — Radent Dental Clinic

"The Leads Are Shit." Every clinic running Meta ads has said it. Here's proof they're wrong.

The leads aren't shit. Manual follow-up, done by humans, is the problem.
Here's how we fixed it — with evidence from 1,567 real calls.

1,567

Calls Processed

25

Bookings by AI

628

Real Conversations

€15–25K

Recovered Revenue

“In the first 10 days, Anna booked 13–14 consultations — nearly double what our receptionist managed in the same period. Three patients had already come in. What really surprised us was that many of these were patients our receptionist couldn't convert. She simply couldn't handle the volume. If the AI keeps improving, it changes everything.”
Dr. Daniel Rapcau

Dr. Daniel Rapcau

Founder & Implantologist · Radent Dental, Arad, Romania

TLDR — Why Read This

More patients from the same ad budget

Without changing your ads, targeting, or budget. The only thing that changes is who picks up the phone.

Stop burning your receptionist on cold calls she hates

No more "I never asked for this," no more hang-ups, no more strangers cursing at your front desk.

See exactly where your money leaks

Evenings, weekends, Monday mornings before you open. This report shows you who's catching them for you.

What happens when every lead gets a callback in seconds

The same leads your team called "garbage" start booking appointments. Evidence from 1,567 real calls inside.

8 min read · Worth it if you spend money on Meta ads.

Is this report for you?

  • You spend €1,000+/month on Meta ads (Facebook or Instagram)
  • Your receptionist says "the leads are garbage"
  • Your cost per booking is climbing, but the calendar stays empty
  • You know leads are coming in — you just suspect most get lost before anyone calls them back

01 — Symptoms

The Monday Morning Problem

You open your laptop Monday morning. Forty-seven new leads from the weekend are sitting in your CRM. You don't feel excited when you see them.

Your stomach tightens.

Because you know what comes next.

The receptionist drops everything and starts dialing strangers. The first three don't pick up. The fourth says "What? I never asked for anything." The fifth hangs up. The sixth wants the price, hears the number, and disappears.

By lead fifteen, she's not qualifying anyone. She's surviving. She's been rejected nine times before lunch and she's mentally checked out.

And those forty-seven leads? By Wednesday they're cold. By Friday, they've already called your competitor — the one who picked up first.

You can't tell which of those 47 leads was ready to spend €5,000 on implants. They all looked the same in the spreadsheet. And your team treated them all the same — which means the €5,000 patient got the same rushed, annoyed, three-rings-and-voicemail treatment as the one who clicked by accident.

47 new leads. Again. I can't do this anymore.

Fig. 1 — Current state of lead management, observed across multiple deployments

02 — Diagnosis

Follow-Up Failure Syndrome

The disease isn't in your ads.

It's in the gap between form submission and the first call.

49%

of Facebook leads never answer the first call

Not because they're bad leads. Because they filled out the form at 11 PM in bed on their phone, and your team called at 10 AM the next day. By then, they've forgotten.

34%

of all inbound calls to the clinic go unanswered

Not because the receptionist is lazy — because she's already on another call. That patient doesn't leave a voicemail. They Google the next clinic.

If your receptionist seems like she's given up, it's probably not laziness — it's from calling strangers who don't want to talk. Sound familiar? Then she starts telling you the leads are garbage. Stops calling after the first attempt. Cost per acquisition climbs. And you start wondering if Facebook ads even work.

Clinical Assessment

The problem is that you're asking a human to do a machine's job — dial 1,256 numbers, absorb 768 no-answers, endure 68 rejections — and still have empathy left for the ones who actually want to come in.

Your receptionist after 50 calls vs. Tellet AI after 1,256 calls Call #1,257. Same energy.

Fig. 2 — Endurance comparison: human operator vs. artificial

03 — Results

1,567 Calls. 21 Days. Field Data.

A European dental implant clinic running Facebook ads for All-on-X procedures (€3,000–€5,000/patient). Leads on leads. Receptionist overwhelmed. Good leads getting lost. They installed a Tellet AI Voice Agent.

1,567

Total Calls

avg 75/day · peak 176

1,256

Outbound (FB follow-up)

every lead called in seconds

272

Inbound

100% answer rate

75

Daily Average

peak: 176 (March 12)

63

Peak Hour

1+ call per minute. Zero quality loss.

Full Outcome Breakdown — All 1,567 Calls

768 No answer / hung up A human gives up after 50. The AI did all 768.
628 Real conversations (>30s) 40% actually talked. Your gold is here.
68 Not interested / rejected So your team doesn't have to hear it.
59 Busy / voicemail Without getting demoralized.
47 Info provided, no booking Seeds planted for future conversion.
34 Callback requested Not a "no." It's a "not yet."
14 Transferred to human The only ones that actually needed a person.
4 Wrong number
25 Bookings made Revenue recovered. By AI. No human intervention.

768 no-answers. 68 rejections. And hidden in the pile: 25 patients worth €15,000–25,000.

The leads weren't garbage. The team was drowning in them.

Want to see these numbers for your clinic?

We'll build a free demo agent with your name, specialty, and schedule.

04 — After-Hours Coverage

38.4% of Calls Happened Outside Business Hours

Of 1,567 calls, 601 were outside office hours. The AI worked 168 hours/week vs. 40 for a human.

298 (19.0%)

Weekend (Sat + Sun)

230 (14.7%)

Evening / night (after 6 PM)

131 (8.4%)

Early morning (before 8 AM)

601 (38.4%)

Total after-hours

"Hmm, dental implants..." *clicks Submit* 11:00 PM Sunday night 5 seconds Hi! I'm calling about your request... 11:00:05 PM

Fig. 3 — Response speed: from form to call in under 5 seconds

Key Discovery Sunday, 6:54 PM — 9 minutes 46 seconds

Patient

"I've lost all my teeth. Dentures top and bottom. The lower one doesn't hold at all. I have diabetes, insulin, history of stroke, and hypertension."

Outcome

Appointment confirmed — Monday 10:45 AM. Details sent via WhatsApp.

Without AI, this €3,000–5,000 patient calls your competitor Monday morning.

05 — Real Conversations

Transcripts from the Field

Case A — Patient Rejected by Another Doctor BOOKED

Patient

"I've lost several teeth. I can't eat. Another doctor said nothing can be done. I'm devastated."

Agent

"I understand, that's discouraging. We frequently treat cases that other clinics have turned away. A free evaluation with the Doctor will determine exactly what can be done."

Patient

"I'd like that, but I'm really afraid of being rejected again."

Agent

"It's completely normal to feel that way, especially after being turned away..."

Evaluation booked — Friday, 4:45 PM. Turned fear of rejection into a confirmed appointment.

Case B — Price Objection Overcome BOOKED

Patient

"It's a lot. I can't afford it."

Agent

"You're not the only one with that concern. We can do the free evaluation so you have a clear picture — no obligation. Would you book, even just informally?"

Patient

"Let's book it."

This is the patient your team would have marked as "not interested."

Case C — Geographic Objection + Dynamic Incentive INTERESTED

Patient

"I'm not from your city. I can't get there."

Agent

"For patients from other cities, we offer a €150 travel bonus per arch, covering transport and accommodation."

No pre-programmed script could have done this — dynamic knowledge base extraction based on the caller's context.

Case D — Emergency Protocol URGENT

Patient

"I had 5 implants placed last week, terrible pain, very swollen..."

Agent

"I understand this is urgent. I'm transferring you immediately. Please stay on the line."

36 seconds. Zero sales attempt. Patient safety protocol followed without exception.

Heard enough? Let's talk.

15 minutes. No commitment. No credit card. No contracts.

06 — Prognosis

Before & After Treatment

Before Tellet:

  • 47 uncontacted leads from the weekend.
  • First 10: no answer. #11: "Stop calling me."
  • By 2 PM: 25 calls made, 1 booking, behind on everything.
  • 200 leads. 8 bookings.
  • €2,000 spent on ads.
  • "Facebook leads don't work."

After Tellet:

  • Monday: 6 pre-qualified evaluations already booked.
  • Each called within seconds. Already spoken to. Already agreed.
  • The phone is no longer the enemy.
  • 200 leads. 25 bookings. Same budget.
  • The doctor asks to increase the ad budget.

07 — Financial Impact

Revenue at Risk

Here's what happens to a typical €2,000/month Meta ad budget without instant follow-up:

Your monthly ad spend €2,000
Never answer
49% → €980 burned
"I never asked for this"
~15% → €300 wasted
Interested, no follow-up
~20% → €400 in the air
Real conversation
~8% → Maybe 2–3 bookings
Cost per booking €400–€700

With Tellet AI

1,256 leads → 628 conversations → 25 bookings in 21 days

At 20% conversion rate:

€15,000–25,000 recovered revenue

Same leads. Same budget. Same campaign. Different operator.

How many €5,000 patients did you lose this month because nobody called them fast enough?

08 — Common Objections

Answered

"Patients will know it's a robot and hang up."

272 people called KNOWING it's AI. 126 became real conversations. People don't care if it's AI — they care that someone picks up.

"Healthcare needs a human touch."

Agreed. The AI booked 25 AND transferred 14 complex cases to humans. Your team handles 14. The AI handles the other 1,553.

"We already have a receptionist."

Perfect. Stop making her cold-call strangers from Facebook. Let her do what she's actually good at. Let the AI do the part she hates.

"Facebook leads are low quality."

The AI found 25 ready patients in 1,256 leads. They weren't low quality. They were unfiltered. Now they're filtered.

09 — Treatment Plan

Your ads are running right now.
Who's calling your next lead?

1

Send us your clinic name on WhatsApp

2

We build a custom demo agent with your specialty and schedule in 24h

3

You hear it handle a real scenario — live, on a 15-minute call

No credit card. No contracts. If you don't see value in 15 minutes, we part as friends.

AI Voice Receptionist for Medical Clinics

Source: European dental implant clinic · 21-day deployment · 1,567 calls · Feb–Mar 2026