Unlike other symptom checkers that rely on static rules or protocols to give you the best guess at a diagnosis, K Health was trained on millions of real anonymized medical records to give you more accurate results that are personalized to your situation.
Hans Tung: You were involved in Wix, Vroom and now K Health. What made you decide to join each of these companies?
Allon Bloch: First, you need to know what motivates me. Harnessing technology to make an impact in society, particularly if the technology favours the consumer, is what gets me excited.
You have to understand that most industries calcify around the legacy incumbents. They focus on their oligopoly, they rake in those high margins. They forget about who their product is serving.
If you look at Wix, Vroom or K Health, there is one unifying thread. We’re using data and technology to completely eviscerate the existing structure.
When Wix started back in 2007/2008, your go-to for building a website as a small business was terrible. You would spend weeks on it, spend tens of thousands of dollars, and it was not scalable. You couldn’t change it without forking out even more. Paying 10 bucks a month, or 20 bucks a month, or five bucks a month to build a website that you had control of was liberating.
Now, K Health is about giving consumers control in their health journey, about what they need to do, and putting the consumer in the driver’s seat.
Aside, I also join places where I can learn in areas where I might have a shortfall. I joined Wix from a consultancy (McKinsey) and didn’t have first-hand knowledge of engineering, product or marketing. I’m a little opportunistic in that sense – I love partnering with product experts and learning from them.
Hans Tung: What are some of the lessons you’ve learnt along the way as a third-time founder?
Allon Bloch: I’ve made so many mistakes that I’ve lost count! It will honestly take the whole podcast.
I’m 50 now. When I look back on my role as a first-time CEO at Wix, my priorities then and now are different. Now, I value life experience and pattern recognition. But above all, the biggest takeaway for me is how to effectively manage people.
People are complicated and technology is easy. How do you manage people? How do you inspire people? How do you get people to care about the company, and deliver great results, not because they get paid, but because they genuinely want the company to be successful for its users.
How do you attract the right people? There’s no playbook for that.
One of the other lessons I learnt is that my biggest competitor is not my competitor. It’s inertia. Competitors often actually provide a lot of value by educating the industry, educating the customers. You have a joint marketing budget. You can educate people to buy cars online and people would go to one or another. Your competition is inertia.
Another key lesson is trying to create the perfect core initial product and creating that mass demand. It’s critical to get that right and get a breakthrough without paying a huge ‘toll fee’ to Google or Facebook. Organizational issues are where by far I’ve made the biggest mistakes.
But that’s also why Wix, Vroom and K Health have become successful – we made more correct decisions than mistakes.
Hans Tung: How did you get ready to be a third-time founder? With serial entrepreneurs, more eyes are on you. There’s more pressure. How were you ready for that?
Allon Bloch: I don’t have a perfect answer to this question. I spent five or six months before jumping in, thinking about the opportunity, talking to people, but not too much such that I got swayed.
Some of the doubts I had – I’m not a physician, what do I even understand about medicine?
So while I talked a little to the people in the industry, I tried to look out for nuances. I predominantly read books and talk to people who are not in the industry and try to get them excited about ideas. I do need to brainstorm with a couple of people. Then you build some kind of theory about what you can do. That’s completely different, and a little bit stupid and crazy. And then I start building a conviction. I typically need a vacation in the middle. And so I was in the Canyons before K Health. After three or four months, I decided that I want to go ahead, and do it. I don’t agonize about the risk of failure. I don’t think I’ll fail. I just think there’s no chance in hell I’ll fail. It’s the equivalent of a person thinking they’ll live forever.
Hans Tung: I have got to ask the most obvious question. You have 2 IPOs, each company worth billions of dollars, one is over $10 billion. Why are you still working?
Allon Bloch: That’s a question that my wife asked me as well. I enjoy it. It’s fun. What would be more fun to try and use resources and technology and work with a lot of smart people to try and solve big societal problems?
I couldn’t think of anything better to do with my time, during my awake hours. I do enjoy being with my three kids and hanging out with my wife and my friends. But this is something that is tremendously exciting. I’ve never woken up in the morning and questioned why I’m doing it. I just spring out of bed and I enjoy doing it.
Robin Li: We’d love to hear the origin story of K Health and how it started.
Allon Bloch: It all really started when I spoke with Adam Singolda, who’s the founder and CEO of Taboola, and also a good friend. We’d always wanted to work together.
We have shared experiences around lack of control around your diagnosis and treatment when you go and see a doctor. When you go to the doctor, they’re sitting with the stethoscope and talking to you with big Latin words. It’s scary. You feel like a little child.
Even if you’re a super successful and sophisticated person, you feel small. You place all your trust on this doctor and follow instructions. You don’t dare or think of challenging him. The alternative is searching via Dr. Google but sometimes that’s even worse with all this misinformation. Control was a huge part of it.
I think it definitely stems from how both Adam and I like to have control – where we eat, how we take our vacations, how we manage money. So, why can’t we control our health?
There’s a personal story behind this, at least for myself. My dad had atrial fibrillation, which I didn’t know anything about. It’s a disorder of the rhythm of the heart. Because of that he was taking Coumadin, which is a blood thinner and the end point is that he had a stroke, because the drug failed. He was not clear how risky his condition was for a stroke. My dad went from being a super healthy 69-year-old who was travelling and working to almost dying. He survived it, but he has speech impediments and was dragging his right arm.
That’s a little bit of backstory. Back to how it started. My co-founders Ran Shaul and Israel Roth, along with a very experienced ER physician Neil Brown, we got together.
We needed a data set big and rich enough to answer the three basic questions humans have when they’re not feeling well: What do I have? What else could it be? How do I treat it? And from there, we could train a machine to do differential diagnosis and treatment.
In the early days, we spent a lot of time defining the core tenets. It needed to have an ability for people to get unbiased information, it needed benchmarks (a dynamic cluster of people that are relevant for your situation), and it needed to be a specific conservation around your health.
As we grew, the founders and myself got smarter about the U.S. healthcare industry and its approach. Medicine is a practice in the healthcare industry. There’s medical protocol, but also everything in the medical universe is generally expensive. We quickly realized the U.S. healthcare system is out of control.
Here in America, we pay probably three times more than what equivalent people in Europe pay! And they live longer in Europe if you do 1-1 comparison. The pre-existing system is the definition of an anti-consumer system that’s just really, really expensive.
That’s where we were coming from. How could we give control and penetrate the healthcare industry, by creating better medicine and a more economical price point? Those are still our guiding principles. But of course, our thinking has evolved quite a lot since then.
Hans Tung: Right, which brings on a very important point. You need data to train AI, to train the system, to come up with a better recommendation. But in healthcare, it is just so hard to get your hands on data. And one thing that you guys do distinguish yourself from Dr. Google and other self-diagnosing apps is you have real anonymized medical records. How do you get your hands on data?
Allon Bloch: We first started looking at people who published articles in science and nature around using electronic medical records (EMRs) for predictive capabilities. We noticed that they typically use claims data, which is what is written for in order to charge insurance. We couldn’t use these as it was very limited.
We also couldn’t get access to raw doctor’s notes – the quantity available just wasn’t large enough. Even so, raw notes need to be anonymized, normalized and you’d need to ETL (extract, transform, load) it.
In my experience, nobody’s done that for primary care and chronic care.
The way we did it was a little fortuitous. There was a structured payer-provider that about two and a quarter million Israelis belong to and they were willing to take a bet on us. They had built this full technology stack with data that was digitized for 20 years. On their side, they were thinking, okay, we have all this data, and we want to partner with companies who are going to create a lot of value from the data.
That was our start.
Hans Tung: What other breakthroughs did you have that helped with access to data?
Allon Bloch: A big part of it was focusing our efforts on primary care conditions – like what a doctor would consider in the first visit. Things like super rare conditions or oncology or cardiology, we left out.
We focused on that kind of mass market. In doing that we were able to build an anthology that’s richer than any other public anthology out there by an order of magnitude. Our machine understands medicine, and understands logically, statistically, and semantically what is having or not having a thyroid. It can understand how a thyroid impacts all the other parts of the body in terms of symptoms and attributes.
It can map all this to a medication, diagnosis and treatment.
We were fortunate to be able to identify the right technology leaders, data science, and engineering. Our CTO, our data science team, they were all great.
Thus we were able to build this initial anthology and classifier with a really small team. It took us just over a year to put this good enough product (or MVP) in the market and get going.
Of course now, we have many data sources, and we also have a lot of our users to use it, so we can learn from our own data.
But at the time, we just had an information layer and no services layer. That came later.
Robin Li: For many years, consumers and the healthcare system were just notoriously resistant to innovation brought by tech. Like you said, your biggest competition is inertia. But when COVID-19 happened, K Health shot up in the charts. Do you see this behavior of consulting with an app and using K Health becoming a norm in the future?
Allon Bloch: Overall I think we’re witnessing massive change. Now hospital groups are telling me that they’re able to conduct things remotely, at times even better than in person.
Obviously a huge part of this change is that people are now tying any symptoms they have to a COVID-19 trigger. They may not be willing to wait a few days to see a doctor physically, they need to know right now if they’re sick or not.
One of the biggest pain points we’re solving is accessibility to doctors – because obviously doctors work certain hours. If the issue can be resolved in a few minutes with an application, if that information can be provided, that becomes very attractive to people. I think people are ready for that.
Another thing to note is the price point. Like I said, healthcare in America is just overpriced. A lot of the stuff that happens in the primary care clinic is wasteful. Multiple people talk to you, someone takes blood pressure and weight, another person draws my blood. That’s a lot of time and money on all these procedures, and who knows if it’s needed?
Here’s the great thing about data – we can actually personalize it. You can create continuity of care. Example – every time you come back from another country, you have a set of symptoms that appear every two or three months. They appear three or four days after it’s here and the app is tracking it. This is how the richness of data lets us identify patterns.
If you’re less well-off and you live somewhere with poor access to medical care, K Health is obviously helpful. But even if you’re privileged – think about this – doctors make mistakes too. Everybody can benefit from richer, better information and improved access to it.
Hans Tung: I think with COVID-19, it makes it a lot easier now for hospitals to work with you than ever before. What were the benefits brought about by the pandemic for K Health?
Allon Bloch: There’s definitely an open mindedness to change because everyone had to move to telemedicine. They needed to do video calls for consultations, appointments.
Things that I thought would happen in 2025 are happening now. Previously the trends in digital health were kind of basic. They weren’t really impactful.
I do sense some movement and pressure from certain hospital groups to continue to innovate.
Bear in mind, data in healthcare is not typically used for health, which is a big statement. It’d be the equivalent to Google saying data of user behavior in search is not going to be used for our search engine, we’ll just use it for ads or billing. Data in healthcare is used for billing. It’s used for workflow.
Very few people access data to say, how do I improve health? But it’s not using systems like K Health. The use of that data is what will separate the true innovators and trendsetters from the rest. For that matter, it’s not just about telemedicine or remote care anymore. I think that’s a foregone conclusion.
We’re looking at the next phase. Have you built something that really makes a difference? So we try to focus when we talk to partners, or partners who kind of intuitively understand that and are not looking just for the basics.
Hans Tung: Help us understand how K is different. What differentiates it from telemedicine?
Allon Bloch: Telemedicine is like a car phone. You remember the car phone in your car. That was innovation, maybe 15 to 20 years ago.
For us, I think we’re taking an approach of reimagining primary care, and information, doctors drugs and tests. It’s not just about the video medium.
Hospital groups and large providers are implementing solutions that are 10 or 15 years old. That’s how they’re approaching things – starting with legacy assets. They’re not thinking from scratch on how to solve a problem – they’re plugging a gap with a basic telemedicine app or system. They’re not thinking about consumers and those needs because they’re saddled with a high cost structure already.
They’re focusing on the wrong questions, which it’s gonna be hard to compete with solutions like ours if they focus on the wrong questions. Even if they have fancy press releases and big store openings.
Robin Li: It’s a good point – doctors are consumers too. They want to make it easy for them. The same way that consumers want health care easy as well. You mentioned in K Health that you can text your doctor. But as far as I know, doctors are not necessarily incentivized to go that route, because the billing codes don’t reimburse doctors for texts, as opposed to, maybe a physical visit. How do we get around that model, especially the way the US healthcare system reimburses people or doctors?
Allon Bloch: Now we’re getting to the complexity of healthcare. I’ll shed some light.
What we don’t do: If a doctor’s visit is about $150 between the insurer and your co-pay, I don’t break my brain thinking of ways to cover that $150.
What we do: How do we build a product that people need? What can we do for our partners in terms of getting distribution and getting paid?
In changing how we think, we’ve fundamentally built a better solution for individuals.
Millennial doctors love working on our platform, millennial users love working on our platform. Plenty of 60 year olds work on our platform too, so we are adding pediatrics. There are whole audiences.
We need to have a full stack, to control that entire experience. I’ve never been good at not controlling that. That’s why we have both information and doctors and treatments. Because healthcare is so big, we will partner with people to be clear, but it needs to be around that same approach of working with consumers. But Hans, I’m just curious.
Hans Tung: How do doctors get reimbursed?
Allon Bloch: We have our own doctors.
We’ve figured out a way to do reimbursements much more efficiently and have the doctors focus just on the diagnosis.
Robin Li: Let’s move on to our quick-fire questions. What is one habit that you have that has changed your life?
Allon Bloch: Nobody’s ever asked me that question. I haven’t asked myself that question. Let me answer it.
During COVID, I really enjoyed running, and I found it meditative. If I’m having a long day of Zoom calls, being able to just go out and run is transformative for me.
When I was in the Israeli military, they used to wake me up at 4am and tell me to put on boots and go running in these heavy boots. I hated it.
During COVID though, I love it. I can’t wait to go and run, which I know sounds crazy. But in that time running, I can really clear my head. I’ve solved two or three things I want to do. I can go back and I’m kind of energized. Is it life changing? Well, in a way. It really helped me manage through the COVID period.