Today we’re talking about Oakwood Dental Practice in the UK. They have 22 dentists, 3 receptionists, and offer a pretty wide range of treatments.
Like any serious clinic, they spend tens of thousands of pounds on advertising. Dental patients can have a strong LTV, but practices still need to keep bringing new people in.
And as often happens in business, if you look a little closer, getting more sales and more customers doesn’t always mean buying even more traffic. Sometimes you just need to fix the “bottlenecks” that already exist in the patient journey from initial interest to actually paying.
When we started looking at the patient journey at Oakwood Dental Practice, we found that they were missing 228 calls over a 28-day period. And no, it wasn’t because they had bad receptionists or a bad team. Part of the problem was simply that some calls came outside working hours, while others came during peak times.
There’s not much point keeping a receptionist at the clinic just to answer calls after 6 PM. And during peak hours, the problem is even wider than that. Sometimes 3-4 calls can come in at the same time, and answering the phone is obviously not the reception team’s only job. They deal with paperwork, help patients who are physically at the clinic, handle all the usual admin work, and of course answer incoming messages as well.
So missed calls are a pretty natural part of running a dental practice. And according to different sources, in some clinics they can reach 25-30%.1
We took the practice’s average numbers and did the maths:
Yes, £90 might not sound like a huge amount. But take your own conversion rate – how many people who call actually become patients – and multiply it by 228 missed calls. Suddenly it doesn’t look so small.
And even more money is made at the next stages of the patient journey. When someone comes in for their first visit and then agrees to a full treatment plan. When they come back again over the next 2, 3 or 5 years. When they recommend the practice to friends or bring their family in for treatment. At that point, the value of a single patient can already run into tens of thousands of pounds – and sometimes much more.
And yes, you’d be right to say that some patients who can’t get through will send you a WhatsApp message or simply call again. But definitely not all of them. Research in the dental market suggests that patients often call 2-3 practices at the same time to compare options and ask questions. So if you don’t answer, there’s a good chance they’ll immediately call one of your competitors – where someone will answer and book them in for their first appointment.
What did we decide to change?
We realised that missed calls were a leak we had to do something about. There was almost nothing to lose here, because the conversion rate on a missed call is always 0%. It’s basically like that person never came to you in the first place.
So we decided to try an AI voice receptionist – Aimée. It’s a specialised voice agent trained on dental clinic conversations and business processes.
We didn’t want to give AI too much control, so we only gave it the calls we were already losing and couldn’t really do much about:
Patients also tend to hang up after around 15 seconds of waiting. So we ended up with a pretty simple compromise: we don’t lose the call, but if a receptionist is available, they still get the first chance to answer it.
Booking · FAQ · Reschedule · Cancel · Transfer · Callback
We listened to the first 20 calls ourselves to understand how the AI handled real patients and whether everything was working the way we expected.
The agent can handle several things:
After the first week of testing, we found another pretty useful feature: the same agent could also be added directly to the clinic website, so patients could talk to Aimée there, ask questions and even book an appointment through it.
The logic was pretty simple. Most patients visit the website from their phone — around 92% in our case — and looking for information on a small screen isn’t always very convenient. So some visitors simply called the agent from the website, asked their questions and even booked an appointment.
What results did we get?
But the nicest part wasn’t even the missed calls we turned into bookings and revenue. The nicest part was the bookings we weren’t expecting at all.
Remember I mentioned earlier that after the first week of the experiment, we decided to add Aimée to the clinic website as well, mainly so it could answer questions?
Well, we forgot to turn off its booking function and CRM access on the website. And as it turned out, that was probably a good thing.
So, the final numbers from the experiment for our practice looked like this:
I think that’s a very solid result – especially considering we didn’t increase the advertising budget at all.But numbers are one thing. It’s much easier to understand what’s actually happening when you see a real call go all the way from the first “hello” to a booking appearing in the CRM.
We recorded a short demo: a patient calls the practice, asks a few questions, chooses a dentist and an available time slot — and Aimée creates the booking directly in the CRM.
How did patients and the team react?
The team
We had to explain to the reception team that the AI wasn’t there to replace them — and that we weren’t planning to fire anyone.
We positioned it more like a junior assistant that takes care of the work they don’t really enjoy doing anyway. That way, they can focus on more important and more interesting work where their experience and attention actually matter.
For the first few days, we listened to every call ourselves. But pretty quickly, everyone got used to it and it just became a normal part of the day.
Patients
The biggest thing we were worried about was a negative reaction from patients. But our logic was pretty simple: it’s better for an AI to help someone than for nobody to pick up the phone at all.
One important thing here — at the beginning of every call, we always made it clear that the patient was speaking to an AI agent, not a real person. I think that helped us avoid a lot of negative reactions.
We also kept the option to transfer the call to a real receptionist. That helped avoid awkward situations too.
Overall, some patients were surprised, but most of them — surprisingly, even to me — reacted pretty normally. They quickly got into the conversation and understood how it worked.
Overall, from both the team’s and the patients’ point of view, the experiment went much more smoothly than we expected. But before drawing conclusions about the money, it’s important to honestly separate what we actually saw in the data from what is still just a calculation.
What do these numbers actually mean?
Here it’s important to separate the real results from the potential economics.
We can clearly see in the practice data that Aimée created 53 bookings in the CRM from calls the team wasn’t able to answer. 45 of those patients actually attended their first appointment, so the £4,050 in first-visit revenue is already a recorded result.
The £32,290 in treatment revenue is different. That’s not money we can already put on the table. It’s a calculation based on the practice’s usual conversion rate from a first visit to an approved treatment plan, plus the average treatment value.
The same applies to the website. We can see bookings that were created with Aimée involved, but without a separate A/B test, it wouldn’t be right to say that the entire increase in website bookings happened only because of the agent.
And that’s exactly why the main takeaway from this case isn’t the specific £4,050 or the 68 bookings.
What about your practice?
For me, the main takeaway from this experiment isn’t that «AI can now answer the phone».
It’s much simpler than that.
If your practice is already spending money on advertising and getting enquiries, before increasing the ad budget it makes sense to look at the last part of the funnel:
How many people are already calling you or visiting your website — but never actually become a booking?At Oakwood Dental Practice, that gap was big enough to generate dozens of extra bookings in just 28 days without increasing the advertising budget.
The numbers could be completely different at another practice. But it’s pretty easy to find out.
See what happens at your practice
Test Aimée on the calls you already lose.
You can run Aimée free for 7 days with 50 minutes of real calls included.
Test Aimée free for 7 days →During the beta, Aimée will also set up your CRM integration for free, as long as your CRM technically supports this type of integration.