AI voice agent for aesthetic clinics: book consultations
How an AI voice agent calls warm aesthetic clinic leads and books consultations with the doctors, with real data from 4,872 calls in six weeks.
## AI voice agent for aesthetic clinics: the short answer An AI voice agent for aesthetic clinics calls people who already know your centre (enquiries, one-time visitors, people who showed interest and never booked) and books a consultation with your doctors during the call. Your front desk stays with the patients in the clinic. In one Italian network of aesthetic medicine centres, DeepAgent made 4,872 calls in six weeks and booked 108 consultations with the doctors. - It works the warm lists your team never has time to call back. - It says it is an AI, explains why it is calling and asks one question at a time. - Questions about prices, treatment results or medical advice go to the doctor at the consultation, which the agent books. - It logs the outcome of every attempt in your CRM, opt-outs included. ## Why warm lists stay unworked in aesthetic centres Many aesthetic centres have a file like this: thousands of people who asked about a treatment, came in once, or filled in a form and then went quiet. Everyone agrees these contacts deserve a call. Nobody has the hours. The people who would make those calls are the same people who greet patients, manage the waiting room and handle check-out. When a patient is at the desk, the call list waits. The contacts are also hard to reach. In the network described below, 7 attempts out of 10 never got to a person. Doing that by hand means many dials for every useful conversation. When someone does pick up, you get a few seconds. In that campaign, 2 conversations out of 3 ended in under 30 seconds. The caller has to say who they are and why they are calling, quickly and the same way every time. This is where follow-up on warm leads gets lost. For the wider clinic picture, read our guide to AI voice agents across the clinic patient journey. ## How the call works, step by step This is the flow used for outbound calls to warm lists. You can adapt it to med spas. 1. The list reaches the agent via API or from your CRM, with the fields it needs: name, centre, original enquiry, consent status. 2. The agent opens with AI disclosure. It introduces itself as an AI assistant calling for your centre, in line with the transparency duty in Article 50 of the EU AI Act. 3. It gives the reason for the call in one sentence, referring to the person's earlier contact with the centre. 4. It asks one question at a time and waits for the answer before moving on. 5. It handles objections. No time: it offers a call back at a better moment. Price: it explains that the doctor gives all details at the consultation. It never quotes prices or discounts. 6. It offers consultation slots with the doctor, taken from your live calendar. 7. It confirms day, time and centre before closing. 8. It logs the outcome: booked, call back, not interested, voicemail, no answer, busy or unreachable. 9. If the person asks not to be called again, it confirms the opt-out and flags the contact so nobody calls again. Here is a short illustrative call for a fictional centre: > Agent: Hello, I'm the AI assistant of the Aurora aesthetic centre. I'm calling because you asked us for information a while ago. Do you have a minute? > > Contact: Briefly, yes. > > Agent: Would you still like to talk it through with one of our doctors? > > Contact: Maybe. How much does it cost? > > Agent: The doctor goes through everything, costs included, after seeing you at the consultation. Shall I book one for you? > > Contact: OK. Late afternoons suit me better. > > Agent: I have Thursday late afternoon or Monday late afternoon. Which do you prefer? > > Contact: Thursday. > > Agent: Booked: Thursday late afternoon, at the centre, with the doctor. You'll get a confirmation. Thank you, goodbye. The agent stays away from treatment details, promises and prices. Its job is to get the person in front of a doctor. ## The case: 108 consultations from warm lists A network of aesthetic medicine centres in Italy (the customer is anonymous) ran DeepAgent on warm contact lists from 1 August to 11 September. Test calls during the July start-up are excluded from these figures. | Metric | Result | |---|---| | Period | 1 August to 11 September (six weeks) | | Calls made | 4,872 | | Conversations | 1,409 (28.9% of calls) | | Consultations booked with the doctors during the call | 108 | | Booking rate | 7.7% of conversations (about 1 every 13) | | Conversation time handled by the agent | 26 hours (1,556 minutes) | | Attempts without a conversation | 3,463 (71.1%) | In that campaign, most conversations were short. Two conversations out of three ended in under 30 seconds. One in four lasted between 30 seconds and 2 minutes. One in twenty went over 2 minutes. The agent handled those 26 hours of conversation, so the front desk never had to. The bookings went into the doctors' calendar from contacts who already knew the centres. The full story, including how the lists were prepared, is in the aesthetic medicine centres case study. ## The contact-rate lever: 7 attempts out of 10 reach no one In this campaign, 71.1% of attempts did not reach a person. The 3,463 attempts break down as follows: | Outcome | Attempts | Share of calls | |---|---|---| | Voicemail | 1,305 | 26.8% | | No answer | 795 | 16.3% | | Busy | 712 | 14.6% | | Unreachable | 591 | 12.1% | | Total without conversation | 3,463 | 71.1% | These are the numbers of one campaign on one set of lists. Use them as a reference point for your own centre. Bookings only happen inside conversations. Once the script works, the contact rate is the first number to move. The customer has planned these tests: - Time slots: which hours of the day get more people on the phone. - Number of attempts: how many tries per contact before stopping. - Recall rules: how long to wait after voicemail, no answer or busy. - The first 30 seconds: a rewrite based on the conversations that ended in a booking, since most calls are decided in that window. A WhatsApp message sent before the call or after a missed answer has been designed. It is not active yet, so none of the results above include it. If your list is older and colder, the approach changes. Our article on reactivating dormant leads with voice AI covers that case. This one is about the people you should call back while they still remember you. ## Setup and privacy The agent needs a list, a calendar and a script from you. Calendar availability comes via API or through your CRM, so the agent offers only real slots with real doctors. Calls run over your telephony through VoIP/SIP or over numbers provided for the campaign. The centre writes the script and the objection answers. You decide how the agent introduces the centre, what it says about the original enquiry and when it offers a call back. Medical content stays out of the script: advice, results, before-and-after promises and prices are left to the doctor. Privacy needs work before the first call. **Check the legal basis for marketing calls with your DPO or legal advisor**, along with permitted calling hours and the rules on promoting medical services in each country where you call. Our guide to GDPR and conversational AI explains what to prepare, including AI disclosure and opt-out handling. If you also run dental practices, the qualification logic is covered in our piece on AI voice agents for dental clinic lead qualification. Other healthcare uses are collected on our medical clinics page. ## What it costs DeepAgent has five plans. Prices exclude VAT. | Plan | Price | |---|---| | Starter | EUR 99/month | | Pro | EUR 299/month | | Growth | EUR 500/month | | Scale | EUR 799/month | | Enterprise | Custom quote | Annual billing saves 17%. Compare what each includes on the DeepAgent pricing page. ## FAQ ### Can an AI voice agent book aesthetic consultations? Yes. In the campaign described here, DeepAgent booked 108 consultations with the doctors during the call, from 1,409 conversations between 1 August and 11 September. The agent offers slots from the centre's calendar, confirms day, time and location, and logs the booking. ### Does the AI agent give treatment prices or medical advice? No. The agent does not give medical advice, treatment results, before-and-after promises, prices or discounts. When a contact asks, it explains that the doctor covers these points at the consultation and offers to book one. ### How many calls does it take to reach a contact? There is no general figure. In this campaign, 4,872 calls produced 1,409 conversations (28.9%), and 71.1% of attempts did not reach a person. The customer is now testing time slots, number of attempts and recall rules to raise that contact rate. ### Does it work with my booking software? DeepAgent reads availability and writes bookings via API or through your CRM. If your booking software has an API or syncs with your CRM, the agent can offer real slots with your doctors. The specific integration is checked during setup. ### Is an AI voice agent for aesthetic clinics GDPR compliant? DeepAgent can be set up to work within GDPR, with AI disclosure at the start of the call, opt-out handling and a log of every attempt. The legal basis for marketing calls, permitted calling hours and the rules on promoting medical services must be checked with your DPO or legal advisor before you start. ## Start with the list you already have Pick one warm list and one centre, agree the script with your doctors and let the agent make the calls your front desk never gets to. Talk to sales → · Create your agent →