VNV Global AB (publ) (VNV) Earnings Call Transcript & Summary

Hi, everyone. This is now seems to be the way to do these events, but it's the first time for us. So I hope everything will work. Welcome to our Capital Market Day. And there was an echo here, but now the echo is gone, so I'll keep on talking. So welcome, and we -- instead of doing one full day of interactions like this, we -- as you know, we have decided to do a presentation from one of our portfolio companies every day this week at 3:00, Stockholm time. So that's the structure. And it's -- obviously, our largest holding is Babylon. So that's the one that starts off this week today and soon kick off with that. And tomorrow, we'll continue with BlaBlaCar, followed by Voi on Wednesday, Gett on Thursday, and then a new-ish company that we haven't talked so much about called SWVL on Friday. So that's the week, and we super much look forward to it. But yes, the only other thing, I think, before we kick off here is that questions will be -- if you want to ask a question, you enter the question into the chat, and then we'll see those questions. And after Ali and Charlie has finished -- after they finished their presentation, we'll pick up some of the questions, and then we'll sort of address the speakers with the questions. And so that seems to be the way to do these things, and I think it will work fine. But without further ado, I'd like to introduce Ali Parsa and Charlie Steel. Ali is the Founder and CEO of Babylon. For some of you who followed us for a couple of years, you've met both Ali and Charlie at our Capital Markets Day in sort of the -- in real-life format, we used to have them in. And maybe, hopefully, we'll have them in again next year. But I think the last one we had was actually at Babylon's offices in London, so that was great. And -- but it's also great to be able to do it in this way. And yes, I think no more introduction needed. So thank you, Ali and Charlie, for joining us. And over to you, Ali.

Thank you very much, Per. And thank you, everybody, for your time today. I will do a quick introduction of what we do just in case for those of you who don't know us, and then we'll delve into some of the more recent stuff we do, for those of you who have been following us, to give you an update of where the company is. We particularly focus a little bit since our Series C and how far we've come from there, if that makes sense, which is the last time that Per and his wonderful organization invested in us. And then Charlie will take you maybe through some of the progress we've made on the revenue front and the growth that we have in front of us. I hope that works. We'll try to keep this as short as possible to give you the majority of the time for any questions you may have during this session. So for those of you who don't know Babylon, I built a chain of hospitals [indiscernible] today, it did very well, today, is the largest chain of hospital in the U.K. And the insight I took from running hospitals is that it fundamentally globally, not just in Britain or Sweden, but globally, we do the healthcare the wrong way down. We wait for people to get sick, and then we deal with that sickness. It's like equivalent of imagine in this crisis, if we did nothing with this pandemic, we just sat back, waited for everybody to get -- to catch a virus and then dealt with them in hospitals after they had significant crisis. Of course, at times of crisis, we never do that because we just simply don't have the resources to deal with the consequences. But almost everything else in health care, we try to do that. So the fundamental insight we had in Babylon, which was found in Babylon, was that if we spent time and effort at the front to give people a lot of health care upfront, we should be able to save the crisis and the emergencies or some of the crises and the emergencies that they struggle with, which happens to be also the most expensive part of health care. So yes, you can imagine if you could park a doctor in your home and that doctor, you could ask any questions you want at any time you want to, the doctor can monitor and observe you and see the trends and be able to predict the trend and help you overcome the worst things that could happen, you will, by definition, have less crisis, less emergencies. That is the fundamental insight behind Babylon. To do that, however, it is impossible. It's impossible to get a doctor in every home in the current model of delivery. So we sat back and we said, how can we make health care accessible, affordable and put it in the hands of every human being on earth? Accessibility today is an easier problem to solve. As long as we can deliver most of the health care most people need on devices most of them already have, that is highly accessible. We do that today in Rwanda, one of the poorest countries in the world, where we have now delivered health care to the entire population. In February, one of the highlights and proudest moments of Babylon was when His Excellency President Kagame signed a 10-year deal with us to provide universal free at the point of delivery care to the entire population of the country. So if today you're in Rwanda, you dial a number even without a smartphone, and you can get to see a clinician or talk to a health care professional within minutes, more than most Western countries or developed countries can get. The problem is there is no accessibility, we talk about affordability. And if you look at where the costs in health care are, they fundamentally sit in 2 buckets. If you cut it by resources, 2/3 of all our health care expenditures go into salaries. Doctors, nurses, health care professionals in every country are among the areas and the relatively most expensive resource of that country. So as long as we continue doing everything on every issue with those health care professionals, there is very little we could do in terms of affordability. Second, of course, is if you cut it by diseases, 70% of all our cost is in predictable, preventable diseases. So in Babylon, what we do is, on one hand, we try to automate as much as possible for human beings to do. The doctors, nurses, clinicians, we try to introduce robotic process management, self-care as much as possible. And on the other hand, so our doctors and nurses can focus on the more high-end and more expensive part of health care, if you wish, a more complex part of health care. And on the other hand, we will try to observe and monitor our patients so that we can catch from them as early as possible the signals of things going wrong so we can deal with it earlier than now. So with that said, we started to say, okay, if I was a patient, how do I look at the issue of health care? From a patient point of view, forget about the system's point of view, so think of the time when Google tried to make information available and accessible to all. They didn't start by saying, What do libraries need? They started by saying how -- What do the individuals need? So we started saying, look, if you wake up in the morning and you have a symptom, that's fundamentally your start with health care. What you really want to know is what is the diagnosis? What's the condition that caused that signal -- that symptom. In reality, that's a matter of curiosity. What you're most interested in is what is the best treatment for that condition. Then you wish for somebody to follow you to make sure you do that treatment correctly. For instance, you [ refuse ] your therapy or take the pills or whatever it is that, that treatment includes. And eventually, unfortunately, sometimes, you do not recover, then you have a chronic condition that needs to be managed, perhaps true life. That influence, symptoms or conditions to treatment to rehabilitation is what your, I call the, health care industry, apart from the fact it's not health care, it's sick care. If you want healthier, we wanted to keep you at the top of your health, what do we do? We start with the health assessment of you. That health assessment will lead to a plan. If you are wealthy enough, you get a coach to manage you through that plan. That coach will monitor you continuously, as we do with our athletes, say. And because we can monitor them, we can predict what goes wrong with them a few days ahead of time. That is your health care and your sick care. And in an ideal situation, you will have a doctor and a health care team that will deliver that to you all the time. Of course, we can't do that with human beings for every person. So what we do is we do it with artificial intelligence as much as possible. Having said that, and it's really important to focus on this, AI in health care is at its infancy. We are in day 1 of developing this work. And I will come back to it, but I think it is important to notice that while you have to view this and we probably do it according to many, including Professor Judea who is the godfather -- one of the godfathers of artificial intelligence and the man who won the Turing prize for AI. For those of you who are not familiar with Turing prize, that's the Nobel Prize for health care, in our most paper that we published on counterfactual predictions, demonstrated how AI can, for the specific primary care conditions, work as well, actually outperformed 76% of the doctors who sat the same test. And his point was that the company has done so well on this. And the new model of looking at health care is counterfactual AI that many machine learning scientists in the Valley -- in the United States may need to rethink their approaches. So on one hand, we're very proud of how much we achieved on this. On the other hand, and I need to be very specific with this, this is really early stages of AI in medicine to be able to do clinically accurate diagnosis, treatment suggestions, so on and so forth. So of course, at some stage, we need to connect people, it's a human being, and we try to do that virtually. And then once in a while, once in 10 times, you need to see people physically. So through this model, we can now not just primary care or basic care, which provide digital-first gateway to all the health services, all members, including proactive management of the chronic condition, a full service to support the pre, post procedures. We just acquired the business in Boston, United States, which will allow patients to go to a hospital and come out. The results, I was just looking at the results in India, where they deliver a lot of procedures and it's fascinating to see that, for instance, infection post surgery by the patients who use day-to-day our business as it's branded in India, it reduces from 3% to 0.06%. And then, of course, we can digitally facilitate care pathway. So this model of AI to virtual to physical has been working really well for us. Of course, I don't want to simplify that. If you look at each aspect of what we do, we have a very detailed pathway for our patients, every step. This is no different slide than what you would see if you go into the offices of Airbnb, for instance. Every step in the patient journey, in our member journey, is taken into minute little steps, and we basically analyze all those interactions. How do we onboard them? How do we use them? And we analyze it for 4 categories. One is the cost of delivery of that step. Two is the clinical outcome in that step. Three is the satisfaction of the member with that step. And four is the scalability of that step. And by basically painstakingly looking at everything and continuously improving on the scores of that little step, you'll end up in an end-to-end solution that works really well. And it does work really well. I share with you now the results of data that we submitted to be published in a peer review paper. I share with you the summary data. This is NHS data that has been provided by -- can we move the slide forward, please? Thank you. The NHS data that has been provided by NHS stores, that data shows that we took an NHS work facility. We increased the number of patients by 30-fold in a 2-year period. At the same time, the patient satisfaction went up to 4.8 -- star, if you want out of 5 star. 93% Net Promoter Score, I mean, the highest international health service in Britain. The clinical quality in those patients went up to the 97 percentile in the country. And at the same time, their ER visits fell by 35%, and the cost was reduced by up to 35%. That just gives you an indication, a proof point that our thesis that if you look after people really well, it will make them cheaper to look after is true. Health care is one of the wonderful industries where better quality cost less, and crisis costs more. Charlie, actually, if you go back one slide, I apologize. I missed this thing around AI. I did talk about this in some detail, but fundamentally building our AI, artificial intelligence, we try to build it the same way that the brain should function. So we created a significant knowledge base, which we add to all the time. That knowledge base has that many of the facts in medicine. We then try to create a language reasoning. We call it a perception. We will add to that feeling perceptions, visual perceptions through time. And we then put that into a reasoning and decision-making engine, we call this a graphical model that is highly auditable, that will allow us to do simulations and planning for our patients, which is the paper I just mentioned to you. It's called the counterfactual thinking, which is called AI with imagination. So instead of saying that this cause relates to this situation, what you do is you think about what-if scenarios. If the following didn't exist, what would have happened? And that allows you to do significantly -- get significantly better results. And of course, we have a layer of learning. So every interaction closes there. So we do today maybe 10,000 AI interactions per day. And every one of those or physical or virtual consultations, doesn't really matter. What matters is the back end of all of this is the same for us. And every time we monitor a patient, every simulation we do, every health assessment we do, it all teaches our brain how to become better. If you think about this, a human doctor only sees about 3,000 -- 7,000 patients a year. A machine that can see today 10,000 patients a day, these numbers are growing fast. Actually, right now, it's more than 10,000. The numbers I'm giving you is probably 6 months old. Then you can see the speed at which the machine can learn. Now maybe to describe our business model, and I'll switch to that and then I'll pass on to Charlie to talk about our growth since we last talked to you. But if you drew a graph of revenue per person versus a scalability in health care, and that kind of talks to the whole of our business model, you will have -- at the beginning, what we do is we give people a fully digital care. So we basically charge our clients for just a provision of licensing of our technology. For that, we can only charge cents per month per member. So a few dollars, say, a year out of the, say, $10,000 in United States that a patient perhaps in average costs, right? What we find at companies like Amwell, who just went public, do this. And you see that Amwell after 10 years has a revenue of around $200 million while it has over 2,000 clients, enterprise clients. And the reason for that is they can only extract such a small amount per patient that they cover. Then we add a layer of personal health assistance. And by the way, on the digital care, today, we collect about a license fee for around 20 million patients across different categories and client bases. Then we have a personal health assistant. These are clinically supervised but not commissions for whom we charge a few dollars a month, if you wish, PMPM, per member per month charges. However, at some stage, we need to connect people to doctors. We do that, that's a business known as telemedicine. People like Teladoc do a great job of it. But that business, on one hand, you collect a bit of money, the same dollars for your digital care business. But on the other hand, you also provide clinical consultations, let's say, $50, $60, $70 a go. And therefore, that increases the revenue you can get per patient. We do that for around 5 million people now globally. And then, of course, you can move on to what we do in [ new chain ] for instance, which is the full provision of primary care and chronic care management. That allows us to go from tens of dollars to hundreds of dollars per patient. There are people like One Medical, who do that in the United States, again just went public. One Medical does it for primary care. It does it in a high-end scalable way because it's primarily based around brick and mortar. And one of the interesting transactions we saw recently was Teladoc's acquisition of Livongo for around $18 billion. Livongo is an organization that has around $300 million to $400 million of revenue. So that was the amount of multiple today. I think it stands at next year's revenue of almost 50-fold. But the reason that was -- and Teladoc gave almost half of its business to Livongo to acquire the business. But the reason they did this was because they could see that they need to access the higher per person revenue and the more valuable part of the revenue. And Livongo does that in a highly scalable way because they use the same model of delivery as we do. Now what we are trying to do is bring all of that to basically what we call Babylon 360 or a full integrated care solution, where we can look after the entire budget of the patient. Oak Street Health, for instance, that went public last month in the United States, does that on a revenue of $800 million. They are now trading at $12 billion. Again -- but they don't take a few dollar per person. They only have 80,000 patients. And with those 80,000 patients, which took them over 10 years to acquire, they have a revenue of $12 million. In U.K., as I said for primary care alone, we have more than that. In the United States, just to give you an idea of the level of scalability we have, this month, we will launch our full scale, what we call Babylon 360 work in the U.S. which is the same work as the integrated care Oak Street is doing. Our launch, our first month, will be with over 65,000 patients. And those numbers are increasing or can increase significantly. We have more news on this coming, which Charlie will share with you. I understand this is a public market event, and therefore, we have to be careful about what we share with you publicly or not. But enough to say because these numbers will be public. With Per and his wonderful team invested in us on our Series C almost a year ago, we're making around $1 million a month, give or take, in revenue. Next month, we will do in excess of around $20 million a month. That's -- of the revenue. That's almost a 20-fold increase in our revenue, and we will probably sign by the end of this year, if everything continues to align, something in the region of $50 million of revenue per month. So that's the level of growth we are seeing in our business. And the reason we can achieve this growth goes back to the scalability I talked about. Having built the foundations, we're now finding that we don't need to set up brick-and-mortar clinic per clinic. We can just turn the switch on and bring in significantly more people and more revenue onto existing platform. Charlie, with that, I will pass on to you as I know you're both more articulate and more intelligent. So you will do a great job.

Not sure about that, Ali, but thank you for the words of comfort. So I think sort of the main thing I want to say here, really, is that not only have we come a long way from even last year, as Ali said, to this year. So we're expecting to see around a 4x increase in revenue in 2019 and 2020. And just also highlight that some type of revenue increase has continued since 2016 as well. But also by the end of this year, we'll have over $600 million of contracts on an annual run rate basis. As Ali said, we've also just signed another 2 contracts, one for around $100 million a year, the other for around $30 million a year, both in the United States, both in value-based care. Those contracts are being delivered as of today. So we started with the delivery of those in the beginning of October. And then, of course, we have the existing business as well, but -- that's continuing to grow. It's been a huge increase from last year. So we'll go into next year, and we're confident at this stage that we'll be going to next year with over $600 million worth of contracts signed. Some of those will take a few months to start the delivery of, which is the reason why you see that the 2021 revenue being below the $600 million. But we're also very confident we can do that again going forward in '21, '22. And how will we do that? So the primary way we're looking to do that is basically delivering more revenues from our existing clients. So one of the contracts that we signed is an acquisition. The other -- is part of an acquisition. The other one is, again, from an existing client. And the other 2 contracts that sort of get to our $55 million or more in December of signed contracts, those are again from existing clients. So one of the things that we've very much done, going back to what Ali was saying earlier, is that the confidence that people are having in the Babylon product and our ability to deliver and scale extremely rapidly is delivering us a lot more contracts. And that was sort of part of the original investment thesis as well is that we continue to upscale from that. Furthermore, we also believe that, again, the existing clients of ours will deliver us even further contracts to get over $1 billion of annual run rate, which we're very confident we can achieve by the end of next year. And as I say, we'll do this from existing clients alone. The split is predominantly U.S. but also quite a lot of revenue coming from Asia as well. That's really our 2 core markets as well as the U.K. as our home market, where we've had a lot of success so far with the NHS, and we'll prove them. But this is only one part of our growth projection. The next part is really so as well as upselling existing clients, we also believe, of course, we can get another new customer. And what we're also finding is that when we get existing clients -- from our existing clients, we can also get a lot of new customers from that as well. So for example, our investment and partnership with major U.S. health insurer has also direct us towards sort of further hospital groups. And again, sort of things like the NHS in the U.K. has got us another validation where we go to upsell and get new clients elsewhere. We will also see though that the M&A is going to be a key part of our overall strategy. And the way that we'll look to do that is consolidating bricks and mortar opportunities in an industry that's basically very fragmented though the use of technology and bring that together to use the scalability of that and our ability to grow and bring more people on to the platform incredibly rapidly to generate both efficiencies and drive further margins but also more revenue. I think the other thing, Ali, I don't know if you want to mention this a little bit, one thing that's happened since the last Capital Markets Day is that Babylon led the Series B for a company called Higi in the United States. Higi has a number of terminals where you can effectively do our health check products in real life. Again, we get more data. We see that as a key way of accessing part of our core market, a Medicaid and Medicare population. And Higi just signed a major contract with Dollar General to roll out its products in a large number of stores across the United States. So we sort of see that as an additional way that Babylon will capture its user base and really get to understand while using it very well in the way that a lot of other companies cannot. That was everything on my deck. I think we can then go into questions.

Thank you very much for a great presentation with lots of interesting detail. And now for everyone listening in, in the chat at the bar, at the lower part of the screen, you can punch in your questions, and we'll read them out. We'll -- if there are many questions, then we'll sort of have to pick some. But I think just to kick us off, my colleague, Björn here is -- will fire away.

So yes, we can start potentially with -- on the U.S. market, which you rolled out beginning of this year. Could you elaborate a little bit on, I guess, the tenders and the process so far by rolling out in the U.S., whether it's different regulatory environment from state to state? Essentially where are you now? Has it been slower or faster than expected? And has COVID and related stuff essentially been good or bad in the pace of this rollout in the different states?

So I think the answer is almost all of the above. Everything you just mentioned, right? So COVID has been dreadful for our society as a whole, everybody, for humanity. There's no good that comes out of this. So I don't want to be selfish about this and said it had some benefits for digital health companies. Whatever benefit it has is irrelevant compared to the costs in which the society had to pay. From our point of view, COVID did a number of things, the pandemic did a number of things, which frankly speeded up where we were going anyway. So Per would know, and we've discussed this many times, that telemedicine, the provision of a doctor behind a mobile phone was already going to become a commodity business. At the end of the day, there's very little technology involved. It's a logistics business. You just get doctors on the other side, barriers to entry very, very low. So frankly, almost anybody can set this up. And I don't say that flippantly because actually what happened after COVID was almost anybody did set it up. So almost every hospital group, every provider group, even single-handed physicians managed to move on to telemedicine and see their patients that away. And those -- and it was a rather scalable business because every consultation needs to be done by a human doctor. So you saw all these telemedicine companies across the globe chasing to find more and more doctors to do the consultation. We had luckily already had told our clients that this is -- we will do that as far as that's the need. So if you go back to that permit in your mind, we will deliver all aspects of the permit if they want just telemedicine or consultation with the doctor on the other side, we will do that for them. But really, our focus was how do we take their entire budget and help them to reduce the cost of delivery while helping their members to stay healthier and looking after them fully by giving them a personal private doctor in their pocket. It's -- and what we see -- so what we saw is, on one hand, all aspects of our business grew. But on the other hand, and that is very important, what we saw also was that the trend we were having, which was getting a lot of new clients almost stopped because if the client wants to replace their existing supplier with us, and we have a better product. We have no doubt that a lot of people go in discussions to do that. So people who want to -- who are basically -- and I'm sure you're seeing this in other parts of your portfolio, whatever you invested, one of the trends I observed is that many new clients became much more difficult. So we all managed to sell more and more to our existing client base and penetrate those deeper. But going to find new clients on a strategic sale, where the client had to make a significant shift, people really need to spend time with you and your team and win confidence. So I think that what we are going to see is a significant pent-up demand for Babylon services that we can then harvest on the other side of the pandemic, which means on this side of the pandemic. And we are trying to prove that our model works better. And having said that, the pandemic has helped that 20-fold growth in our revenue to come from it. I hope that gives you an answer which says we both saw acceleration of our revenue in places that we didn't expect, but we also saw delay in getting revenue from some places that we expected it to come.

Fantastic. Thank you. Now I think that covered Björn's question, and now you've really gotten everyone going. So there's a -- I'll fire away some questions from the people who are listening in. And I think the first one that came out, and this is for you, Charlie, is to talk -- you're talking about the potential MRR of $125 million. And what would be the time line on that?

Yes. So if I think about the MRR of the $55 million, so that is at the end of this year, I think as was clear on the slide. The MRR getting to $1 billion-plus, we hope to land most of those contracts by the end of next year.

Charlie, may I clarify, just so that there's no confusion. When you say that the MRR of $50 million at the end of this year, you mean signed contract?

Signed, yes.

Not delivered and booked contracts. That will take a few more months to roll out.

Yes.

Thank you. There are several questions on the company's funding needs, and if those will involve an IPO or listed -- as a listed company, if that gives any other benefits from -- than just cash rates.

Ali, do you want me to take this? I'm happy to.

This is a question that can get us into trouble. So I absolutely [indiscernible].

So I think on the -- so to sum it off, would an IPO help M&A, I think absolutely. But obviously, it gives us cash if we do primary within that IPO, but also it then gives you by the second valuation, which then means that doing acquisition using equity for the [ company ] will be easier. I think also the other thing that it also does is gives more options to other instruments, say, debt and convertibles because you have a back-end clear takeout, which you can do using your equity or something else. So I think when you look in the market about sort of some of the funding that people have raised to date, away from Babylon for a second, a lot of people made a lot of recent convertibles, for example. And they've been hugely popular in the public markets. And I think from our perspective, we see that as a much easier thing to do, very low cost of capital when you are a public business than when you're a private one.

Thank you. The next sort of -- there's several questions, which sort of asked you to comment on what bottlenecks you see to growth. Yes. What do you need to unlock this growth?

So we don't really need a lot. There is -- one needs to be careful, right? So we just did a 4x growth on annual revenue, as I said, the 20x growth on monthly revenue. And we're continuing to, as you saw from Charlie's projections, to see significant growth. But we need to also be careful that we digest every feast before we go back to take more. I think in health care or in any business, you need to think about, are you building a company for the next 20 years? Or are you building it for the next 2 years? And we have always built it for the next 20 years. So if you actually look at the history of Babylon, there was a period in the first 3 to 4 years of our existence where we almost had 0 revenue because we invested heavily in building the foundation on which we can then grow the revenue. And so for us, the growth that we projected, it looks like we can get quite a lot more. I mean, look, let's just put this in perspective. A single client of ours, Centene Corporation has over $20 billion of revenue -- sorry, over $100 billion of revenue, sorry, my apologies. 20 million members, over $100 billion of revenue. Single digits of percentage of that revenue coming to us to manage those patients for you because remember our pitch to them, our pitch is you have no control over these patients. Most of them are Medicaid patients. They basically have often not even a primary care physician. So they just wait for a crisis and go to ER -- I mean hospitals. As you know, ER costs a lot of money. And one visit can cost more than their entire annual fee that they collect for that member. So when we can then take that member and show them and manage that member for them, the saving could be substantial for our partner and for us. So one single client can make us grow 10-fold from where even the numbers Charlie shared with you. We just need to make sure we digest this in a manner in which we can sustainably grow over a long period of time. The promise we made to Per when he invested in our business was that we are in one of the -- we are in the world's largest industry in health care in a $10 trillion industry. And there is no reason why somebody can't build a $1 trillion business in health care in the next 20 years. So we hope that is us. If it's not us, it will be somebody else. For humanity, it doesn't really make a difference as long as somebody solves this problem. But if it is going to be whoever it is, you need to play this to win in the long term rather than get greedy in the short term.

Great. I think sort of the follow-up from the audience on that is it's sort of alongside if you need the management bandwidth, data, employees overall. And I think there's some questions also, what's your headcount today? And what will it be 3 years out? How do you see that sort of -- that part of your capacity pool or OpEx?

So I'll take the first one, and Charlie will take the second one. You're absolutely, right. And it's a very wise question that really the only strain on any company is our ability to attract talent. So great people who have seen that kind of growth and have had the operational knowledge of how to manage it are not that many in the world. If you think about how many companies in the world move at that speed and how many of the executives have been through that journey, they usually have made enough money that they want to retire. How many of them are prepared to come back and do this again? You see that human resources is our biggest challenge and management capacity capability. So we're constantly building on that. But it's a very wise question. And it is our significant -- my significant focus is how do you continue attracting very good talent to Babylon. Having said that, we recently actually -- we used -- we took advantage of the situation in COVID in terms of sitting back and saying that if we were going to build, if crisis were going to happen, especially in the first few months, nobody knew how this is going to plan out. And we said that if we wanted to be steady ahead of the time and try to use this in order to significantly make ourselves more sustainable from a cost point of view, what do we have to do? So we managed to do actually significant cost-cutting in the business and reduce some numbers, which invariably have an effect on our ability to execute all of our plan. But we can actually -- with less numbers and less [ volume ] still execute on most of our plans. But Charlie, I'll give it to you to...

Yes. I think from that, what I'd also say is we're starting to see the operational leverage of the business coming through a lot more as well. So clearly, as we scale, we hire more people, add sort of more clinicians, more people to operate in our operation centers, but most of them are going through cost of sales. When we look at the below-the-line costs, we're actually seeing some of that staying, relatively speaking, quite flat year-on-year, particularly also going forward. The other thing I also highlight is that -- is a key differentiator is Babylon 360 from a financial perspective is that we get assigned to these patients upfront. So in order to get the revenue, we do not need to do a load of marketing, which any piece of service business does need to do. So from our perspective, we also see that as a key financial competitor advantage, looking at the Babylon 360 model, is that there's a lot less marketing involvement as well. So a lot less marketing cost.

Thank you. A couple of questions on the U.S. Is the model you're running a fully capitated one, so are the elements on the fee-for-service? And then also, if one could elaborate on what kind of segments commercial versus Medicare and Medicaid.

So we have 2 what we call flywheels. On one hand, we provide clinical services. That's our first flywheel. And then this -- on the other hand, we use the technology that we have to use, and we license that to others. So our business model on clinical services is that we do all of the pyramid. So if somebody wants a fee-for-service, we sell that to them. For instance, in California, we deliver fee-for-service solutions for Centene for about 2 million, 2.5 million Medicaid patients. And in -- but we also do a capitated model, where we take the entire health care budget. And we look after that group well. And as I said, we are cautious with that. So we're launching with about 60,000 people in U.S. now to make sure that works before we expand. And that's in Medicaid primarily and Medicare. But look, we are agnostic. That's a very U.S.-specific way of paying for it. So if you're an employee, your employer pays for it. If you're elderly, Medicare pays for it. If you are poor and in need, Medicaid pays for it. In U.K., if you think, we have all of the above. We have the elderly, we have the young, we have the very poor, we have corporate members. And all of them are either paid for by the NHS or by their employers. So we are agnostic from the payer's point of view. In U.S., we've already had our Medicare deal. We already have our Medicaid deals. We will shortly, hopefully, do a corporate deal. So we will experiment and see the peculiarities of each. We will build our platform or software to be able to deal with the nuances of each because they're each different in payments, in eligibility, all of that. And once all of that is ready, then we will roll it out at speed.

Thank you. And there's a question on who you see as your closest competitor? And what are your key differentiating factors that allows you to win about -- over -- yes, win the deal.

In every segment of that pyramid, have many competitors. When it comes to the top of the pyramid where we really want to be, there really isn't anybody who does end-to-end digital base care. So Oak Street Health, which as I just said, Iora Health, [ Charlie Med ], these are all very good companies who do a great job and actually achieved very good cost savings and very high quality of care. But the good brick and mortar, they just basically set up a clinic, look after patients, very small number of patients, it takes them a very long time to grow. We can do what they do at scale through a digital-first model. Just to give you an example, we were shortly in the United States, announced a deal we did. You started with something like 15,000 people. The partner we did it with have tenfold as many patients to do this. If they were going to do that with one of these brick-and-mortar people, then for them to go from 15,000 to 150,000, for instance, they would have had to build a clinic at the average cost of $1 million a clinic then get doctors and nurses to go and reside in the little towns that those clinics are in. And then bit by bit, sell those clinics, lose money during the infrastructure of that until they set up to 2,000 people each. We don't have any of these costs. We basically just increased the number of people, if you want, in our call centers and the number of relations we have with the existing clinics to see that 1 in 10 times that people need to be seen physically. So that, I would say, scalabilities are key advantage. Now there are others who are trying to do what we do through acquisitions. So they're basically saying, oh, we're in just in telemedicine. We're going to buy chronic conditions. We're going to do this. We're going to do that, and we're going to get there. It depends on who you are and what you feel. Reaching that the reason Tesla is a much better car than a General Motors car that wants to be electric is because everything in a Tesla works together. Tesla does 90% of its own software. Every piece is built to work together. If you acquire and patch things, capabilities together, first, we will find most of these capabilities are not integrated. I think I read in the recent filing for one of the telemedicine companies who've done a lot of acquisitions that all their previous acquisitions are still not integrated. I'm not a big believer in patching many acquisitions together and thinking the worst. I actually know of no company -- you are investors, so you will have a much better knowledge than I do. But I know of no company who's built a substantial global business by just acquisitions. Amazon, Tesla, Apple, all these companies were built on a basis of a very solid foundation. And that's what we're trying to do and grow organically with some acquisitions, but those acquisitions are marginal on the side to add to either capability or a segment entry.

Thank you. One more question in the U.S. before this one back home in the U.K., and that's on Mount Sinai. And we announced a rollout in -- with Mount Sinai, and how is that developing? And if you're seeing revenues for that contract?

Mount Sinai was something that we did very quickly during COVID in order to provide Mount Sinai with a relief, so that not everybody goes into the hospitals, that they have the capability of doing either consultations or just self-management through the -- through our app, which worked very well during the crisis in New York. Frankly, in the confidence of this room, this is a bit of a distraction for us. New York is a -- if you go to -- we don't want to really double down on this fee-for-service model. If you just think about it, let's just take a number, $50 million a month revenue that we try to target to do. At $50 a consultation, that's 1 million consultations a month you have to do in order to get to $50 million of consultation. To put this in perspective, if we just took patients at $5,000 annually, we need -- or $15,000 annually, depending on which part they are. We need far, far less patients to look after. And considering a patient in average needs -- will have maybe one interaction with us every other month, just a number of interactions will be a fraction of the million for the same revenue. And more importantly, we can manage their health care better, so we can deliver a much better service than just the telemedicine service. So I think in business, and I'm sure you appreciate that, you're defined as much by what you don't do as you are by what you do. So we're already there to help our community for what they need at the time of their need. That is really important for us to focus on what we think we can win at rather than try to do what others could also do.

Thank you. We have time for a couple more questions. But one question is if you could elaborate a little bit around the 35% cost reduction in the NHS?

Those data is NHS data. Those are not our data. We're publishing all this, as I said, in a peer review paper. We're just waiting for the publications to tell us when is the time of the publication because these things take time once you do peer review papers. But I actually don't think that should come as any surprise. I mean, you just think of your own family. I had a doctor that was open 24/7 and was available to you all the time, you just use doctors a lot less or emergency and crisis a lot less. And in health care, all the cost are united in the same class. So you take somebody with mental health issues, the way the NHS or most health systems deal with it is that they allow this person to have crisis and then they end up in emergency. Maybe unfortunately they self-harm themselves or maybe they have an episode of -- a critical episode, and they end up in emergency, which costs a lot more. Often, all these people need is a crisis management of their condition. This is what we try to provide. So unlimited access to our doctors, to our professionals, which allowed them not to get to that expensive crisis. Let me put these in numbers for you. A consultation, let's say, cost us GBP 30 a person. If we do a hundred consultations, that is 2 consultations a week for somebody with a mental health issue. And that cost us GBP 3,000, let's put it that way. If that person has one single episode of crisis, which will lead to them to go into an emergency and be hospitalized in a high observation situation, that single episode per night will cost GBP 3,000 a night. So you can just see how much you can take out of the system by looking at the people proactively rather than waiting for their crisis.

Thank you. So I think we are soon running out of time here, but if we do 2 more questions. One question is if it's possible to sort of segment your clients? And what percentage of your customers use all the services in the pyramid versus customers who use 1 or 2?

Well, actually, right now, all of -- most of our customers, the fact that we could grow our revenue base so much is because we upsell the same -- the further up of the pyramid to existing customers. So basically, most of the customers are our existing customers, who will buy in the lower part -- on the upper part. We're incredibly excited to be able to get out there and now sell to new customers, which is the vast majority of people out there are mobile customers, exactly the same. And almost all of our customers are now -- these are, say, key customers are buying across the pyramid floor. And most of the revenue comes from the top.

Thank you. So the final question here, it's a good ending question, and that's what type of operating margins would you consider are achievable for Babylon?

So I think that our margins on software will be like every other software margin, around 70%. Our margin on clinical services will be around 30%, 35%. And our margins on capitated model we hope will be between 25% to 35%. But 35% of $10,000 is $3,500. 70% of $10 is $7. So we are a cash business, and we much rather take the cash than we are worried about these things.

Crystal clear. Thank you very much, Ali and Charlie, for spending time with us and being so generous with your time and also your -- being so generously being open on these questions and the level of detail that you provided.

Thank you. And if I can one more time, thank you, Per and your team, for all your wonderful support throughout the years. It takes a lot of patience to help a company to do something fundamentally right, and you've always done so. So I'm grateful to you and your investors. So thank you.

Thank you very much. Thank you, everybody, and please join us tomorrow.