3 weeks ago
April 1, 2021
Babylon Health’s Michela Simonelli on Digital Content and App Localization
Babylon Health Content Design and Localization Lead, Michela Simonelli, joins SlatorPod to talk about hyper-local digital content, localizing high-touch health materials, and the growing telehealth market.
Michela shares her professional journey prior to Babylon Health, which began with studying audiovisual translation, subtitling, dubbing, and re-speaking, and saw her spend 11 years “agency-side” in multiple language service providers (LSPs). She describes her transition to the user-side with Babylon Health as “massive.”
Michela discusses how content design interacts with localization at the digital health provider and collaborating across functions with epidemiologists, internationalization teams, data scientists, and many others.
She also broaches the subject of machine translation (MT), explains why there is currently no interpreting provision for medical appointments, and identifies an opportunity for LSPs that can provide a holistic service for telehealth.
First up, Florian and Esther discuss the language industry news of the week, which saw two M&A and funding stories in the media localization space.
Private equity-backed Iyuno completed its acquisition of major rival SDI Media to become Iyuno-SDI Group, the largest pure-play media localization company globally, with combined revenues of around USD 400m in 2020. Iyuno Founder and Chairman David Lee steps into the role of CEO of the combined group, while former SDI Media CEO Mark Howorth takes on the mantle of President.
A few days later, UK-based ZOO Digital raised GBP 7.4m (USD 10.3m) in a share placement on the London Stock Exchange. The publicly-traded company said it plans to spend the proceeds on building hubs in India and Southeast Asia, expanding international business development, and on “longer range” research and development, hoping to grow revenues 2.5x to USD 100m in the mid-term.
Florian shares news from multilingual video conference startup KUDO, which secured an additional USD 21m in a Series A round, announced this week, bringing total funds raised to USD 28m.
On the flipside, he talks about a 250-respondent survey conducted by the Belgian Chamber of Translators and Interpreters (CBTI-BKVT), which revealed that conference interpreters in the country saw their income decline by about 50% following the onset of the Covid-19 pandemic.
Stream Slator webinars, workshops, and conferences on the Slator Video-on-Demand channel.
SlatorCon Remote May 2021 | $130
A rich online conference which brings together our research and network of industry leaders.
Florian: You are working with Babylon Health as their Content Design and Localization Lead in the Operations team. Can you just tell us a little bit more about Babylon Health, what the company does and also a bit about your personal background?
Michela: I work for Babylon Health. It is a health tech company. It is a digital health service that combines AI with clinical services, so you can check your symptoms on the app or the web-based part of the service or you can book an appointment with the GP. It is all done virtually. We also work with the government here in the UK so if you want your GP to be Babylon, you can do that and then you can also have face-to-face appointments with the GP. It is a digital slash in-person healthcare service.
As for me, I have been in London for almost 13 years this year. I am originally from Italy but my entire career has been in London. I started with a master’s degree in audiovisual translation, it was one of the very first degrees that they did. There were about one or two books that you could buy actually on subtitling at the time and now it is definitely changed. We studied dubbing, voiceover, and subtitling. We also do respeaking, if you are watching the BBC news and you have live subtitles, they usually come with a respeaking technique, which is quite interesting. At the time it was very new. There were very few courses anywhere, even across London, I think there were two or three universities offering that.
I started as an intern at a small company and then worked on the agency side for the last 11 years until I joined Babylon Health last year. It has been just over a year. I joined as a Localization Lead and now I have expanded my role on the content design parts of the operations. Content design is the people who take the content that has been written and make it better, let us say, for the patients in our case, because those are our customers. Then that content is what moves on to localization so it was a natural expansion of my role to look at the end to end process of the content creation.
Florian: How was the transition from the vendor side to the client slash user side? Was that a big change or did it feel fairly natural?
Michela: It was a massive change. The way that you work is very different so you have clients, but they are your colleagues. When you need something, you need to go to your colleagues for the whole creation of how the product is, the ideation, the purpose, the business strategy. All the information that you never get when you are on the buyer slash agency side and you always wish you would get to do your job better. It is definitely very different. Personally, it was a very positive change.
Esther: I am interested to hear more about the content design element of your role. You said it was a natural progression, but can you explain that a bit more and how it flows into the localization element as well?
Michela: When you are on the agency side you say, I wish I was asked this question before, so now we will not have to face this problem of changing this image in this video because it is not appropriate for this market. When you are on the client side or the buyer side, the situation changes slightly, but you still need to work for it. The information does not just show up on your desk, you still need to go and ask the questions and make sure that you are involved. With the majority of the organizations that I have seen, localization falls under the content department because it is content at the end of the day.
I have found myself working closely with colleagues from the content team and there have been a few projects that the team have taken on that related to how the content team was designing the structure of the sentence and how it would affect the localization. We naturally started working more closely together to improve how that part of the work was done so that by the time it reaches us and we have to deploy this content to 13 different countries, then we do not have to go back and do the work again. They needed some support on certain things so I am overseeing the processes and the prioritization of projects for the entire chapter.
Esther: Can you give us a bit of an update in terms of what kind of content you are localizing, the number of languages, territories, anything you can share on volumes?
Michela: Currently I think we are live in 13 languages. That includes language variance or what we would call Canadian English or US English. The type of content is mainly focused at the moment on the content that is within the app so all the patient facing content. In terms of volume, the last time I counted the entire product so if you want to use all the features of Babylon, it is just over half a million words. It is potentially quite a large piece of work when you go to a different market. We do not have every product in every market, but that is the extent to which it can go. It is quite high.
Florian: What are some of the challenges around the specific medical content? You mentioned that you work with vendors and in-house doctors to validate the medical accuracy of that content. How does that work? Is there a final review step? Do they login to some system?
Michela: From the language localization side, we use a TMS that the doctors have access to. It is quite complex because you have the medical content, the regulated content and you have the safety to take into account for the patients when you are localizing the content. There are so many different people that contribute to this process. It goes a bit further than localizing or transcreating or adapting because there is a lot of work that gets done even before the content gets to us.
An example that I always make is, your head is hurting and the diagnostic around your head hurting can have billions of associations with different things. If you are in the UK, for example, then maybe it can be associated with the flu or something else. If you are somewhere else in the world and you have a fever, it might be malaria. It is not a choice that we make from a language perspective, but you have a whole team of epidemiologists. If you are delivering this content in Malaysia or in Rwanda, what can it be connected to? There is this whole work that happens even before the content arrives to us, to localize which is very complex and very fascinating as far I am concerned. In Rwanda, we work with the government so Babylon is like the NHS for the UK. We provide health services to the entire population of Rwanda.
Then the doctors get involved at the end as well so after we have localized everything, we always use people in the country because it is very important to have people who actually live there. They can spot things like if the app tells you to call an ambulance, is it something that you will do in that country or are ambulances available to you, is this a thing? We want people who live in the country to be able to spot anything that might need changing and after that process is done, it goes back to the doctor’s in-country to review it and sign off. It is quite long, but it does not feel like it when you do it because it is business as usual. It is definitely a thorough process.
Esther: What kind of technology are you using to support your content and localization? Is it proprietary? Third-party? What kind of tools are you using?
Michela: We developed some tools in-house especially for the integration with the AI tools so wherever, we call it the brain, the doctor brings forward. There are some tools that call out to other tools that we use so we do have a TMS, which is not built in-house. It is off the shelf and then we have tools in-house that connect to it to help call the strings and deploy them.
Florian: When you are primarily localizing for apps, are there any particular challenges there and any differences like iOS versus Android? Is there anything noteworthy there?
Michela: It is like these other things that you learn when you start working on the buyer’s side or how many levels of complexity there are to something that you do localize. The good thing is that yes, there are restrictions, the screen size in which the user is going to see. The app affects, the length of the string of the content, so there are a lot of things to take into account, and these are usually picked up at a testing step done before deployment. The good thing is that there are teams in-house that build for those platforms, the internationalization. Actually, the code associated with the string is done by people who do that. It is taken into account. It is a limitation sometimes but luckily, there are specialists who can definitely support that.
Esther: Where does machine translation fit into your workflow? If at all?
Michela: I have explained the end to end process because we do not really translate or localize per such, it is almost as if we are creating a whole new set of content specific to the region. It is also such a regulated market, it is very difficult. It will not be very beneficial for us to do machine translation. We would not benefit from just doing the post-editing because we will end up having to change so much anyway. Something that I considered when I joined Babylon because you are not as familiar with the content, is immediately you think it is software localization, so surely machine translation fits well into this. Then you start seeing all the different needs and things that you need to do. At least as things stand now, it is not something that we would benefit from but it is not to say that maybe tomorrow we will look at what is new in the industry.
Esther: You said there is no interpreting at the moment. Is that a strategic decision? Why is that? Do you think you could use technology if you did want to introduce something along those lines?
Michela: It is not a conversation that we have had. There is a lot of work to do and you just go and start doing that. At the same time, it is not a conversation that we are not having. We do not do it at the moment. I am speaking for myself, not on behalf of the company, because I have not really seen the data. I think the demographic that we serve right now in the UK, for example, is younger, 30, 45 demographic and mainly English speakers. That might be a reason why. Also, whenever we go to other regions, our clinical services are already offered by clinicians in those regions so it is already in the language of the country so it is not a need per se. The only scalable aspects of the app would be if we wanted to go to a country and offer the service in different languages.
In Rwanda they Kinyarwanda. We do have translators that we were working with for a while who were very good and specialized in medical content. I do not know if I have been lucky enough and whoever came before me worked hard to find them, but so far we have been quite lucky and we have been able to translate it.
Esther: What is the process? What is your setup with your current language service providers and how do you onboard them? How do you find them?
Michela: We use language service providers so we do not have direct freelancers. We recently ran an RFP for localization. RFP requirements, in terms of how we work with them, is we ask them to work on the platform that we use. We need that because medical devices need to have a record of changes, who worked on what so working on the same platform means that we can have a record of all of that. The advantage of having been on the LSP side is to be able to anticipate what they might need to do the job the way that you want it done. It is very much a collaboration and I am sure that anyone who works on the buyer side shares the same perspective. It is best to over explain, make them a part of it, if there is a product change or a strategy change, explain it, have regular meetings, talk to them, make sure they have everything that they need to do the job. I always make the effort to get them as involved as possible in the process.
Florian: What is the rationale around the decision to not go directly to freelancer, but still have an LSP involved? What is the value you are getting from that intermediary LSP?
Michela: One of the reasons is having somebody we can train once and they can take the responsibility to pass on all the knowledge. We set up a process and they follow it so, otherwise, we would probably end up having 10, 20 project managers in-house to manage all that. The second one is scalability because if we were launching into a new country, then you would have to use all our internal resources to recruit, onboard and all of that. Whereas when you have an agency, they can do that for you. You just make sure you show your requirements so you manage to scale more quickly.
Florian: What about the impact of Covid-19, the business, the content?
Michela: I think Covid-19 validated our business model. Which is to have the digital remote consultation with clinicians and it definitely sparked the interest in the app. That is more for commercial conversations. In terms of what it meant for us on the localization side, the entire company had to gather and create a new part of a product that was specific to Covid-19. What that meant for localization was that we had to update this content that was being created for all the regions where we are every day because of the regulations. Every single country has different regulations, wear a mask, do not wear a mask. Whenever anyone came to our app to ask for advice we had a team of doctors keeping it up-to-date and we had to regularly update it. There was quite a lot more work for us from Covid-19, from the pandemic. The positive is that people have started using telehealth more and appreciate the value of it, which means that the future can only look bright for a business like Babylon.
Esther: How do you think the telehealth space is evolving, other than it is growing? What is next for telehealth or more specifically localization within telehealth?
Michela: I was thinking about this and as we were doing the RFP for a localization vendor, what I found is we had a lot of the agencies come with a presentation describing how they do software localization, how they do clinical localization, and how they do medical device localization. Apart from one, nobody has really looked at this as a digital telehealth service all together and I just think there is an opportunity there for agencies to look at it. You have asked me about interpreting so the moment we request it, it would be good to see an offering that says, we know that you have an app and you have doctors, but we have this platform where you can have interpreters and we are going to make it very quick for you and scale what you offer as opposed to doing it in silos. There is an opportunity there for localization agencies to offer a more holistic approach when they look at medical content, as opposed to looking at it in software or clinical or medical devices and as separate entities because if you look at Babylon, it is everything together. There is an opportunity and more and more people, governments are now moving to us to a telehealth approach.
Florian: In terms of Babylon’s own trajectory, what are the plans? What do you have in store? What are the next markets?
Michela: At the moment we are expanding quite a bit in the US, in Missouri particularly. We are offering Babylon 360, which is kind of an end to end healthcare. If you ever listen to Ali Parsa, the CEO, talk, he would very well describe the difference between healthcare and sick care. Actually, if you look at the NHS or the government’s healthcare systems in general, they tend to look after sick care. When you are sick, they look after you. Whereas, the company strategy is to look more at preventing sick care and focusing on health care. If you have a chronic disease, like diabetes, it gives you the tools to look after yourself. That is how the app is restructuring slightly to offer more support on health care. For the moment we have started in the US. I am not too sure where we are going in the future, but definitely, the mission is to make healthcare accessible to every human on earth. Hopefully, we will get there.