How to adapt your marketing to the online behaviour of medical travellers


The decision making of modern medical travel patient has been influenced, like any other business ‘entity’, by technological globalisation. The prospective customer invests time finding the best treatment for themselves based on a variety of factors including for example cost/benefit, return on investment (ROI) and perceived risk.  While the numbers may not be the same, but their decision-making process is very similar to the behaviour of a corporation.

Let’s take an example of a Croatian dental clinic trying to attract more medical travelers from the UK within a given marketing budget.

1.    Medical tourists use a predictable search pattern

When people are looking for a solution to their health problem, they usually start with an online search via one of the main search engines. Google is the dominant search engine in Europe (Google’s search market share in UK is close to 90 percent).

Measuring the conversions (in this dental travel case, a ‘conversion’ was an action, like a sign-up or a phone call, when a mail was sent), we find out that the search behaviour of UK users is very similar to searchers worldwide.  People start their journey describing their problem using more general, low ROI search terms (e.g. fixed dentures, full denture cost) but end with search terms that partly describe the solution (i.e. implants Croatia, dentist Croatia). 

2.    Search engines offer optimal timing

Search engines, by default, have the best timing of all advertising platforms.  Your message is displayed in front of the searcher right at the moment they are searching for an answer. No other advertising or messaging platform has similar abilities.

Furthermore, your message can be extremely relevant to the search query because the searcher is using specific keywords. The keywords trigger the ads only if the search query is close to the keyword.

Therefore, the message is highly relevant to the searcher and is displayed at the right moment.  Look, for example, at what comes up when I Google ‘laser eye surgery’:

3.    How can I use machine learning to improve conversion rates?

While there are lots of benefits in analysing the behaviour patterns of prospective customers, and using the right timing for advertising, there is still one problem - what is the best possible message I can use to attract patients?

Free to download API software, Google Cloud Natural Language, offers some very helpful ‘machine learning’ analysis. Machine learning is a subset of artificial intelligence that often uses statistical techniques to give computers the ability to "learn" with data, without being explicitly programmed. Via some powerful machine learning models this Google Cloud API analyses the structure and meaning of subjective text posted online.  You can use it to understand what people think (positive, negative or neutral – also known as their ‘sentiment’) of your product or service, from the mentions in text documents, email, blog posts, messaging apps, or even from the recorded conversations happening in a call centre.

To understand it better, let’s look at these two ad messages and their respective cost per conversion:

The first advert has almost a 2.5 times higher cost per conversion than the second ad. By using machine learning to look at people’s ‘sentiment’ of the two ads, we can start to work out why.

Sentiment analysis of ad 1:

Sentiment analysis of ad number 2:

Looking at the overall scores, it’s clear that the sentiment of the first ad is 4 times lower than the second.

After testing a lot of ads, we have find out that almost every ad with a higher sentiment score has a lower cost per conversion.

Even social posts with a higher sentiment score have better engagement rates, and emails have better response rates once opened (the opening rate of an email depends mainly on the recognition of the sender and the strength of the subject line, but once opened the response correlates positively with the sentiment score).

If only for this, I’d suggest it’s worth testing your mails, web content, posts or advertising messages against the sentiment score.

4.    Why else should you consider machine learning?

Big sets of online data contain a lot of different signals (peoples opinions are quite naturally multiple and varied). Machine learning analysis is really the only way to accurately and efficiently process this data to help you work out the most efficient message.

For sure the results will be subjective, but once you’ve invested some time in getting to grips with this machine learning world, you’ll start appreciating it’s value (just like the so called ‘IKEA effect’ - when people have to invest personal effort in a product, it’s more worth to them).

And then beyond the sentiment score, you can measure all sorts of other interesting and funny things (salience, magnitude etc.).

At least for now, this tool is free - so why not give it a try?

Miroslav Varga is PCC Quality Control Manager at Escape, Croatia and is a search engine and online advertising specialist with in-depth knowledge of Google’s tools, especially Google AdWords.  He presented on this topic at the 2018 IMTJ Medical Travel Summit in Athens.  Have a look at the Summit film for an overview of the event.


Marketing to attract international patients for medical tourism

Resources, 25 April, 2017

Miroslav Varga, Escape, IMTJ Medical Travel Summit



Do you have an article that you’d like to share with the medical travel industry?

Publish for FREE on IMTJ.


Related Articles

UK missing out on the Gulf market?

20 November, 2019

Don’t give up on government-sponsored medical travellers from the Gulf

Dental tourism to Costa Rica

20 November, 2019

Cost and quality drive medical travel to Costa Rica

China boosts inbound medical tourism offer

06 November, 2019

New policies push China's Hainan pilot zone

Setting up a medical travel business

23 October, 2019

Lessons from the front line

Promoting Malaysia’s medical travel sector

23 October, 2019

MHTC promotes Malaysia healthcare to the UK