One dental practice. Found in 14 days,
from recordings they already had.
Then I scored 25 more practices on the same framework.
Only one thing was consistently done well: answering the phone.
Two independent methods.
14 days of the practice's own inbound call recordings, scored call by call against a fixed 7-stage framework. Every call where a ready patient did not book was counted, at the practice's own average case value.
400 inbound calls per month, ~100 of them new patients. A conservative 10-point lift in booking rate, at the practice's stated average case value of £1,850.
Both landed in the same range. I'm publishing the lower one.
I scored 25 practices on the same 7-stage framework, 0–5 per stage.
Every call was a genuine new-patient enquiry about implant treatment.
The rubric, with every anchor written out.
| Stage | Average | Scored 1 or 0 |
|---|---|---|
| 1. Greeting and contact | 3.4/5 | 4% |
| 2. Taking control of the call | 1.3/5 | 72% |
| 3. Understanding the patient | 1.6/5 | 52% |
| 4. Presenting the practice | 2.2/5 | 20% |
| 5. Handling price | 1.4/5 | 68% |
| 6. Pre-close | 1.5/5 | 60% |
| 7. Asking for the appointment | 2.0/5 | 44% |
The phone is answered well. Then the call is abandoned.
Estimated booking probability, as the calls actually ran:
median 10%. 89% of calls sat at 20% or below.
Estimated probability had the same call been run properly:
median 45%.
Sample: 25 practices, single market, average case value above £1,800. Scored by one analyst against a fixed rubric.
Then I called seven London practices myself. Same enquiry, same afternoon, single implant. Greeting 3.1, everything after it 1.9 — and five of the seven ended the call without taking my name, number or email. The seven London calls, with extracts.
Patient asked about a single implant. The receptionist knew every number cold — brands, components, staged pricing. She was accurate throughout.
The patient was comparing three clinics. He was quoted £1,215 and booked elsewhere. Nobody asked which tooth. Nobody mentioned the surgeon. Nobody offered a second opinion on his scan.
Patient's tooth had not been extracted yet. That makes him a candidate for immediate implant placement — one surgery instead of two, three months saved.
He was never told this.
Three surnames. No experience, no specialism, no reason to prefer any of them. To the patient they were three strangers.
The clearest example in the sample. The receptionist introduced herself properly, named the practice, was warm and quick with the price.
Then she asked no questions at all. Not one.
Estimated chance this patient booked: 5–15%.
Estimated chance had the call been run properly: 50–60%.
None of these were bad receptionists.
They were polite on every call, accurate on every price, and pleasant to deal with. That was never the problem.
A US dental consultant with 37 years in the industry described the identical pattern to me, unprompted, before seeing any of this data: practices "just wing it," and once a caller doesn't book on the first call, there is no follow-up at all.
Different market. Same failure.
A report is a snapshot. The calls in this document were recoverable — in most of them the patient was still deciding for another 48 hours. By the time anyone reviewed the recording, they had booked somewhere else.
Stopping it needs something that reviews every call within minutes and tells the person who answered what to say when they call back. That's what I build. But the number comes first.
Review 14 days of inbound call recordings from one of your practices. Score every new-patient call against the same 7-stage framework. Return a written breakdown with the calls that were lost, what was said, and what each one was worth.
The recordings, and your average case value. Nothing else.
The breakdown within 7 days, and a number.
Nothing. If the number doesn't justify the time it took you to send the files, I'll say so and go away.
You reply to this email. That's the whole first step.
Getting the recordings out of your phone system is my problem, not yours. Most systems export in a few clicks — I'll walk your practice manager through it in one email, or get on a ten-minute call with them and do it together. You'll get a private upload link. Nothing needs to go through email.
On the recordings themselves: they're deleted once the audit is delivered. No patient names appear in the report — callers are referenced by time and reason for calling. If your DPO wants a processing agreement signed before anything moves, send it over and I'll sign it.
If you'd rather not share recordings at all: say so, and I'll call one of your practices myself as a patient and send you that breakdown instead. Same framework, nothing required from you.
Run this on one of my practices