Why do medical tourism patient enquiries get lost? From the first message to treatment with a CRM and AI
A message from Germany arrives at 02:14 and waits on a coordinator’s phone until morning. We explain where patient enquiries get lost in medical tourism, and how a CRM and an AI assistant change that same night.
Without a CRM (The coordinator’s phone): A message that arrives at night waits on a phone until morning, then two different quotes go out.
- 02:14, Lena K.: Hallo, ich interessiere mich für Zahnimplantate. Können Sie mir Informationen schicken? (Translation: Hello, I’m interested in dental implants. Could you send me some information?)
- Unread
- Source: ?
- 7 hours later
- 09:40, Ayşe: Quote, Implant treatment, €2,400
- 09:41, Burak: Quote, Implant treatment, €3,150
- 09:52, Lena K.: Ich habe schon eine andere Klinik gefunden. (Translation: I’ve already found another clinic.)
- Lost
With a CRM (Clinic · Single inbox): The assistant replies at once, and the coordinator takes over with a summary in the morning.
- In working hours, the assistant steps in if there is no reply within 10 min
- 02:14, Lena K.: Hallo, ich interessiere mich für Zahnimplantate. Können Sie mir Informationen schicken? (Translation: Hello, I’m interested in dental implants. Could you send me some information?)
- Channel: WhatsApp
- Language: German
- Source: Ad in Germany
- Outside working hours: the assistant replies at once
- 02:14, Assistant: Vielen Dank für Ihre Nachricht. Um welche Behandlung geht es genau, wann möchten Sie kommen und haben Sie ein Röntgenbild? (Translation: Thank you very much for your message. Which treatment are you interested in exactly, when would you like to come, and do you have an X-ray?)
- 02:14, Assistant: Den Behandlungsplan und den Preis schickt Ihnen Ihre Koordinatorin nach der ärztlichen Beurteilung. (Translation: Your coordinator will send you the treatment plan and the price after the doctor’s assessment.)
- Explicit consent given
- 02:19, Lena K.: Zwei Implantate, im März. Das Röntgenbild schicke ich mit. (Translation: Two implants, in March. I’m sending the X-ray too. (Roentgenbild.jpg))
- Working hours have started
- 09:05, Coordinator took over: Elif (German). Assistant’s summary: Treatment 2 implants, Preferred month March, X-ray Sent
- Treatment · Hotel · Transfer in one record
In medical tourism, patient enquiries are lost not only because of the ad budget or the website design; they also get lost in the first hours after the message arrives. A WhatsApp message that comes in at night waits on someone’s personal phone, two people send the same patient two different quotes in the morning, and no one knows which ad brought the patient.
In this article we tell the story of the same night twice: first without a CRM, then with a well-built CRM and an AI assistant. Then we set out the assistant’s limits, the eight jobs a CRM should do in medical tourism and the questions to ask about health data. We covered bringing patients to your site in our guide to attracting patients from abroad, and the site itself in our health tourism website guide.
A night without a CRM: where enquiries get lost
It is 02:14. A patient in Germany writes to the clinic’s WhatsApp line: they are interested in dental implants and would like some information. The line is open on a coordinator’s phone, the coordinator is asleep and the message waits unread until morning. When no reply comes, the patient also fills in the consultation form on the website.
At 09:40 two coordinators reply to the same patient separately. One saw the WhatsApp message, the other saw the email from the form. Each sends a quote with the price they know, and the patient receives two different figures from the same clinic. The reply that comes shortly afterwards is clear: the patient has found another clinic.
No single mistake loses the enquiry on this night; small gaps add up:
- The message stays on a personal phone: If the coordinator is off, asleep or with another patient, no one sees the enquiry.
- The same patient gets two quotes: Because the channels cannot see each other, two people deal with the same patient separately.
- The source is unknown: No one recorded which ad or search brought the patient, so budget decisions are made by guesswork.
- Follow-up is forgotten: There is no system to remind the team when to write again to a patient who has not replied.
The same night with a CRM
Now let’s send the same message to a CRM built for medical tourism. It is 02:14 again, and the patient writes in German again. This time the message lands not on a personal phone but in a single inbox that gathers all of the clinic’s channels.
- A record opens: The message is recorded with its language (German), its channel (WhatsApp) and its source (the ad campaign in Germany).
- The record is assigned: It goes to the German-speaking coordinator, and the other coordinators can see who has the patient.
- The assistant replies at once: Because it is outside working hours, the AI assistant replies in German and asks the patient which treatment they have in mind, when they would like to come and whether they have an X-ray.
- The coordinator takes over: In the morning the coordinator takes over the conversation with the assistant’s summary; the treatment, the preferred month and the X-ray that came in are clear at a glance.
- Follow-up is planned: If the patient does not reply, a reminder of when to write again is added to the record.
- The programme is built in the same record: After the doctor’s assessment, treatment dates, the hotel and the airport transfer are planned in the same patient record.
The difference between the two nights is not how hard the team works but where the enquiry lands. A single inbox, a clear assignment rule and an assistant that gives the first reply at night settle in advance who does what in the morning.
What an AI assistant does and does not do
A clinic AI assistant on the WhatsApp line does not replace the coordinator; it leaves the coordinator a conversation that is ready to take over. In healthcare, what the assistant does and does not do should be set out from the start, in writing.
What the assistant should do:
- Reply only with information the clinic has approved: the treatment process, the documents needed, accommodation and transfer options
- Collect the information the coordinator needs: the treatment in mind, suitable dates, and X-rays or photos with the patient’s consent
- Pass the conversation to the team whenever it is not sure
What the assistant should not do:
- Make a diagnosis or recommend a treatment
- Give a final price: the coordinator sends the plan after the doctor has assessed it
- Make any promise the clinic has not approved
The clinic also decides when the assistant steps in. The starting rule we suggest is simple: outside working hours the assistant replies at once, and during working hours it steps in if the coordinator has not replied within 10 minutes. We set up assistants like this as part of our AI and automation service, and define together with the clinic’s team when the conversation is handed over to them.
The 8 jobs a CRM should do in medical tourism
The short answer to what a medical tourism CRM is: a patient tracking system that carries an enquiry in one record from the first message to treatment. Here an enquiry does not move through sales steps alone; it moves together with language, the doctor’s assessment, documents, accommodation and transfers. When you choose or build a system for CRM and ERP, check whether it does these eight jobs.
The same enquiry in a CRM panel
- Inbox: Lena K. (WhatsApp · German), Daniel R. (Instagram · English), Sanne V. (Web form · Dutch), Claire M. (Email · French)
- Lena K.: Elif · German, Automatic assignment: language
- Patient journey: New → In contact → Quote → Appointment
- Quote versions: Quote v1 Invalid, Quote v2 Valid
- Patient record: Treatment 2 implants · March, Hotel Scheduled, Airport transfer Scheduled, Documents X-ray, Explicit consent given
- Tomorrow 10:00 · Follow-up: Lena K.
- Sources: Ad in Germany (longest bar), Google search, Instagram, Website
1. A single inbox
The web form, WhatsApp, Instagram and email should land in the same inbox. Whichever channel the patient writes from, the whole conversation should appear in one record and on one timeline.
2. Assignment by language and coordinator
The moment an enquiry arrives, its language should be recognised and the enquiry assigned to a coordinator who speaks that language. With clear assignment rules no enquiry is left without an owner, and two people never deal with the same patient separately.
3. Follow-up reminders
A patient who does not reply, or who asks for time to think, should not be forgotten. The CRM should link the date of the next contact to the record and remind the coordinator in charge when the time comes.
4. Quote versions
After the doctor’s assessment the plan may change and the quote may be renewed. Every patient should have one valid quote, with older versions kept in the record with their dates, so the patient never receives two different figures.
5. Photos and documents
X-rays, photos and test results should sit in the patient’s record, not on personal phones or in email attachments. Every document the doctor needs for the assessment should open from one place.
6. Appointments, accommodation and transfers in one record
Treatment and follow-up appointments, hotel days and the airport transfer should be planned in the same patient record. When the programme changes, everyone on the team should see the current version.
7. Ad source tracking
For every enquiry, the campaign, the country and the language it came from should be recorded from the first moment. Without this information the ad budget is split by clicks, not by the patients who arrive.
8. Reporting
Reports that show the enquiry, initial assessment, quote, confirmed date and treatment steps by channel and by coordinator show you where you lose patients. Use reports to find the gaps, not to police the team.
We compared whether to cover these jobs by adapting an off-the-shelf CRM or with a system built for you in our guide to a general CRM or custom software.
Test your own clinic
The easiest way to see where you stand today is to write to your own clinic as a patient would. Send a test enquiry once at night outside working hours and once during working hours, and note the following:
- How long the first reply took.
- Whether the reply was in the patient’s language and answered the question.
- Whether the reply asked for what the coordinator will need: the treatment, the dates, the documents.
- How many people got back to you when you wrote through the form and WhatsApp at the same time.
- Whether the channel and the campaign the enquiry came from were recorded.
Run the test from a different number and in the language of your target country, and do not tell your team in advance. Your notes show which gap to close first.
Health data and KVKK: what to ask when choosing a CRM
Article 6 of the Personal Data Protection Law No. 6698 (KVKK) counts health data as special category personal data. Because X-rays, photos and messages about treatment will be stored in the CRM, choosing a system goes beyond a technical decision. When you choose or build one, ask these questions:
- Who sees which record: Coordinators, doctors and accounts do not need access to the same information; permissions should be separable by role.
- Is access logged: You should be able to see which user opened which record, and when.
- Where are the explicit consent texts: It should be clear which text the patient is shown in the form, on WhatsApp and in the assistant’s conversation, and how their consent is stored.
- Where do the data and the AI service run: On which servers, and in which country, the CRM’s data and the conversations processed by the assistant’s AI service are kept.
- Can you export your data: If you change systems, you should be able to take your patient records and conversations with you in a readable format.
This section is not legal advice. Decide the privacy notices and consent texts, the conditions for processing data and the services you will use together with a legal adviser.
Our experience
Tuedion is a brand of Ege Bilgi Teknolojileri, founded in Aydın in 2010.
We describe the structure we set up for health tourism businesses on our patient tracking CRM for health tourism page: enquiries from the web form, WhatsApp and Instagram land in one patient tracking panel, and every patient can get a quick reply in their own language. Treatment dates, the hotel and the airport transfer are planned in the same patient record. We set up and manage multilingual ad campaigns in target countries and connect the enquiries from the ads to the same panel.
We set up the record system in line with KVKK: authorisation, access logs and explicit consent texts are part of the structure. The AI assistant in this article, for its part, is something we set up as part of our AI and automation service: it replies only with information the business has approved and passes the conversation to the team when it is not sure, and we define the conditions for handing over together with the business.
The software we developed for health tourism has been in use since 2023.
Frequently asked questions
Does a small clinic need a CRM?
Look at the number of channels and people rather than the number of enquiries. If enquiries come through more than one channel and more than one person replies, a small clinic can also send two quotes to the same patient or leave a message without an owner. A single inbox, assignment and follow-up reminders are often enough to start with; you can add the other jobs as the need arises.
Can an AI assistant give a patient a price?
It should not give a final price. In medical tourism the plan becomes clear after the doctor’s assessment; the assistant tells the patient this openly and collects the information needed. The coordinator sends the plan and the price after the doctor has assessed it.
How do WhatsApp messages reach the CRM?
The clinic’s WhatsApp line is connected to the CRM through the API that WhatsApp offers businesses. Incoming messages land in the single inbox, and the coordinator’s reply goes to the patient from the same number. This way the conversation sits not on a personal phone but in the patient record that the authorised team can see.
How is patient data protected in a CRM?
With role-based permissions, each user sees only the records they need for their work, and access is logged. Documents are stored inside the record, not in email attachments or on personal phones, and are backed up regularly. Decide which data is visible to whom, the consent texts and the services to use together with your legal adviser.
If you want to look together at where your enquiries land today and draw up a plan for a single inbox, assignment rules and an assistant, let’s talk about your project. Preparing the notes from the test enquiry you sent your own clinic before the meeting makes it easier to turn the plan into concrete steps.