About this episode
Konstantin started Kaia Health after he paid $50,000 to attend an elite treatment center for his back pain.
They've since helped over 625,000 users and received $125 million in funding. But this isn't his first rodeo. In 2015, foodora, his food delivery startup, was acquired by Delivery Hero.
So a big thread here is that a lot of success in the entrepreneurial journey is about the macro forces rather than the micro. It's about being there in the right market at the right time.
In this conversation
- He paid $50,000 for a four-week residential pain rehab program, realized "everything we do there you can just do in an app" — and turned it into a company that's since helped 625,000 users and raised $125M.
- The contrarian tech bet that won: while better-funded competitors spent six years insisting wearables were the answer, Kaia used only the phone's camera and computer vision to correct exercises like a real physical therapist — and the rivals eventually capitulated and switched.
- His counterintuitive read on clinical risk in digital health: "there's no way a clinical trial comes back with bad results" — because when patients actually do the exercises, outcomes are so strong that "if you replaced the word app with medication, they'd say it's the best drug I've ever seen."
- Why the US is effectively the only market where digital health has a real business model — and why the expensive part of building Kaia was never the AI or the trials, but the go-to-market.
- A serial founder's theory of success as timing, not genius: sell a food-delivery startup that grew 20% a week, then meet at Oktoberfest and start a pain company nobody believed in — plus his take on the "hero license" and why the education system trains great people out of building.
Transcript AI-generated
So you fast forward, you exit from that, and then you think your next challenge is going to be musculoskeletal health, back pain, that kind of thing. How did that come about?
Well, I just had terrible back pain. I saw like 100 specialists in New York, got crazy healthcare bills. I saw that the number one cause of personal bankruptcy in the US is healthcare bills, and I was like, that's so weird — coming from Europe, where healthcare is pretty much free compared to the US at least, I was really shocked. So you're sick, and you're bankrupt because you're sick? That didn't make any sense to me. And the other thing that didn't make sense is I thought back pain is a relatively easy condition. I didn't really think it's so difficult to get rid of.
Then I saw that one third of Americans — which is 100 million — suffer from some pain problem, and worldwide more than a billion suffer from some pain. I'm like, that's insane. It's not like cancer, where you need some biotech innovation that a Nobel Prize winner has to do.
So walk me through what happened at the rehab center.
“I ended up at a pain rehab center for four weeks, full time. There was no medication, no surgery. It was $50,000. And everything we do there, you can just do in an app. And that's what we did.”
Konstantin
In the end, nothing helped. I really tried pretty much everything except surgery, because I read that only 10% of surgeries show improvement. So I ended up at a pain rehab center for four weeks, full time. There was no medication, no surgery. I was like, this is so stupid, that this is really hard to access. It was $50,000. And everything we do there, you can just do in an app. And that's what we did.
So what learnings do you take from the foodora journey into health tech? What were the things you did differently that someone without your experience might not have known about?
I think the main thing was that at Kaia I trusted myself a lot more. At foodora in the beginning, I always thought other people knew it better. But what I did do right at foodora is I talked to the restaurants and our users a lot. I delivered food to our users myself and gave feedback on the driver app. I ordered food a lot myself too, and gave feedback on the packaging, on everything. So I think I did that right, by instinct maybe. But then at Kaia, I just knew how to do things, so I didn't have to think about how to do things anymore.
How to hire people, how to assess people, how to manage people — managing people was completely new for me. I managed a lot of people in a very short time, and I made all the mistakes I think you can make when managing people. Right out of university you have no clue how to manage people, because you're barely managing yourself.
That helped — to have more confidence in how to do things. But obviously Kaia is so different from what we did. At least I knew better how to structure the company, which people are really good, which people just look good on a CV but then actually don't work well.
Did you have an anxiety in the back of your mind — look, I'm a young guy, I've had success with foodora, the market was good, the timing was good, but now I'm going to enter this new industry that I don't know or don't have a lot of experience in. Maybe I was a one-hit pony. Did you have that thought or not?
Well, I understood the problem extremely well, because it was the number one problem in my life for two years. If you ask yourself — or those who listen — did you ever have one problem that was on top of your mind for two years? I don't know if that's the case so often. I really knew it, and I did all the different treatments, so I understood that problem from the user's perspective extremely well. And I think that's the only thing you have to know about what you're doing. Because what I see a lot is people start — what's the hot topic? Oh, let's start something in that topic. I won't say AI now, but AI. But then you don't actually know if that's a real problem.
Because I had this crazy pain for two years, I knew that was an insane problem. I'm a pretty happy person, but in hindsight I'd say I had a medium clinical depression, right? Because there's pain 24/7. I go to a bar, I want to talk to you, but I only think about the pain, because it's dominating my brain — like a really loud sound that you have to listen to and you can't hear anything else.
The other reason I was never worried is because our idea was just to digitize that therapy and put the app in the app store for free, and then just do the next, more commercially focused thing. So we started with no pressure. We just said, let's do this because it seems interesting. But we didn't really think it would be such an interesting market.
Wait, so does that mean with your initial product, you just thought, let's just make this, put it out with no expectation, really?
Yeah, and then we just wanted to give it to a small team to run it and have B2C subscribers and all of that. We wanted to make it more like a nonprofit thing. And then we found out how healthcare works. So we changed that over the years a little bit.
And then the whole digital pain management area became so hot. Who would have thought? When we started it was the most unsuccessful thing — people were like, what are you doing? You should do something in logistics or in food. Pain is kind of the worst thing to do. No one did anything related to it. We went to the app store, and the only app there was from a pharma company that gave you three exercises and then wanted to sell your medication.
Well, I mean, if you can get a pain solution right, pain is the ultimate inelastic demand, right? If you're in pain, you'll give your last dollar to try and secure relief.
Yeah, but it looked like there was no market, because the healthcare system doesn't work like that. It doesn't allow an app to just take care of things. I basically wanted to have it for myself, because the pain kept coming back. I wanted this app so I could do this whenever I have pain, and then just make it accessible to everyone who wants to treat pain like I do — at home, and not always schedule a meeting or go four weeks into rehab. So it wasn't very ambitious in the beginning. That's the funny thing.
Let me give you a typical story of a patient who comes to see their primary care physician with pain, and then maybe you can tell me how you thought you could fix it using an app. That's part of the disbelief of this, right? So a patient comes in, they've had this chronic pain, you try medication, that doesn't work — and you can't use a lot of the more effective stuff because it's addictive or has a side effect profile. You try physiotherapy, and the adherence is pretty poor, and even when people do it, it may or may not help. And then, as you pointed out, surgery doesn't have very good outcomes either. So you're kind of stuck. You've got this person with this pain, and there's a cycle where they're becoming less active because they're in pain all the time, and they're unhappy, and it's this negative spiral in their life. Not a lot you can do. It makes no sense to me how a digital app would be the solution when even doctors struggle to treat it. So how did you think about that?
Yeah, I think the answer is that doctors don't treat the pain. They diagnose it, and then they send you to a PT or to a pain management center — and there are not a lot of pain management centers. Most people have never seen one, because there are like 100 in the US or something.
And the thing you have to do every day is physical exercises, psychological exercises, and medical education. For all three of these things you need a therapist. Therapist time is very expensive. If you get a PT session, you never do more than 15, maybe 20, minutes of exercises. But you have to go there, you have to go back, you have to schedule it, you have to change into workout clothes somewhere. There's a lot involved, so you don't do it that often, because it's a financial constraint and it's not convenient. In the busy lives we live, I don't know who can take one and a half hours for one PT session. And often it's in the middle of the afternoon. For me, it was so stressful to even make that time.
So the therapist's job — we've now digitized in a pretty interesting way with technology. We did studies showing how well we correct your movements compared to a real therapist, and showed that we're as good as a real therapist. And that's what I love about the business: we use technology — computer vision, voice AI — and then show that it increases the quality of care. That's kind of the wow experience in our product. People are like, I thought I needed a therapist. It's sometimes good to go see a therapist face to face, but most people can't do that all the time. And they need to do something every day. So they hang around and use our product.
I want to come back to how effective and how great what you've built is. But assuming it's great and it works, the biggest problem you'd still have in your initial product is adherence, right? We know that even taking pills — if someone has blood pressure, diabetes, a long-term condition — it's something like 50% adherence. And for therapy I guess that's even worse: doing these annoying exercises in the middle of the afternoon, no one's got time for that. So how did you tackle that part? Was there anything smart you did to get people to actually do the exercises?
Well, we thought that if we very quickly improve someone's pain, then the motivation is high enough to do it. Funnily enough, in Germany that's a lot more the case than in the US. In the US, we play around a lot with other incentives. Something that works really well in the US is financially incentivizing the users: if you do 10 therapy sessions, you get a certain amount of either just dollars or a reduction in your premium, or whatever it is. And we also add coaches who build a relationship with the users, so you feel more accountable as a user.
The whole therapy is then more personalized. This is our daily thing now — how do we increase activation and retention? And we ask a lot from the user. We don't just say, hey, order food now. We say, roll out your exercise mat, get on it, do these exercises. So the investment we ask from the user is a lot higher. Activation and retention is a real thing. And don't forget — if you go to a PT face to face, almost no one does the exercises at home. So that's basically what we have to compare ourselves against: getting from that 1% of people who continue doing the exercises at home to a much higher retention, and then showing cost savings for the health plans and insurances we work with.
Can you give me a bird's-eye founder view of when you're building a digital health solution — which markets look attractive? You've obviously had incredible success in Germany, and you've entered the US as well, and both happen to be two of the largest healthcare systems in the world. Is that the metric you go after, or are there other things that make something more or less attractive?
“The US is pretty much the only market right now where digital health has a clear business model — which is why the only big digital health companies are in the US.”
Konstantin
I would say we didn't have incredible success in Germany. In Germany, we just helped shape that industry, or get it started. But we still don't really make money in Germany.
In the US, that digital health industry is 10 years older. The health plans know how to onboard a member base, and there's already a defined price tag that people are used to paying, which allows you to build a profitable company. In Germany, all the digital health startups, I think, are struggling to build a business, because the prices are negotiated down a lot. That destroys the unit economics — because in Germany we're prescription-based, since that's the law. So it's a lot of effort to sell to the doctors. It's like a pharma company. And doctors aren't so used to prescribing apps yet, so we basically have to pay a doctor sales force to go after physicians and convince them to do something. That's always expensive.
So I think the US is pretty much the only market right now where digital health has a clear business model — which is why the only big digital health companies are in the US, or have the US as their main target market.
Can you fill me in a bit more? So you're at Kaia and you're like, OK, we've built something pretty cool, we've studied it, it's effective, and now we need someone to pay for this. What does that journey look like — that part of you being a salesperson? How did you think about that, and how did you get reimbursed?
So we started just B2C, subscription-based, and we had 10,000 B2C subscribers. Then we realized patients are used to their health plans paying for medical things. And also the VCs believed more in the B2B route, and we were raising from VCs. It also seemed like all the big healthcare companies that were already successful were focusing on B2B. So we were like, OK, let's not innovate everything. Let's just innovate pain management, but not the go-to-market.
Then we went after the health plans and the self-insured employers in the US and just showed them how great our product works. Our USP is that we use the camera of the phone to correct your exercises in real time with computer vision. Nobody had that back then. We were featured at the Apple Developer Conference a couple of times, worked with their computer vision team — that was really cool. So in the end, we're a computer vision company. We just look like we use that technology to treat pain, because I identified it as the most valuable use case in my life. And luckily, that's the case for a lot of other lives too.
So your magic sauce is that you use computer vision. You don't need any wearables or any other devices. You're just literally using the camera on the phone.
Yeah. And we have competitors in the US, and they always have been using wearables. They raised a lot more money than us — and we raised quite a lot of money. So for years they insisted that wearables were better. But guess what? Wearables only track the point on your body where you wear it. So what they were able to do is count repetitions, and that's not a very high-value use case. We were able to correct exercises, like a real PT would do if she was standing next to you.
And so after six years of them insisting wearables are better, they finally shifted the strategy completely to computer vision. But now they're obviously in year one or two of trying to catch up, and it's really hard. It's not an easy problem. So that was nice, because we were betting on the technology, and everyone — including investors — was like, oh man, it looks like it's not the winning strategy. Look at the wearable strategy, and they raised more money.
But the users — again, we listened to the users, and they were like, this is so nice, I would never put wearables on my body. So we just asked our users, and they said, it's so convenient, I just put the phone on the floor, I don't need to learn how to use an additional piece of hardware. And my co-founder and I, we don't believe in hardware. We believe Apple and Samsung and whoever else — they're great at developing hardware, and we should just use the technology that's available. Because if any new sensors become available, these devices will have them, because they're such gigantic companies. They'll have to leverage whatever powerful technology comes out. That was a very high-conviction belief we had, and it paid off.
So why have you raised so much money?
Because the thing is, if you go B2B, it's really expensive in the US to acquire B2B customers. That's the main reason. And we're in Europe and in the US — we're one of the few companies, I think the only digital health company, that successfully operates in Germany and the US, which on the other hand adds some complexity that makes it a little bit expensive. We didn't want to give up Germany, so we invested a lot in Germany, but it starts to pay off, hopefully soon. We didn't give up.
And if you look at what the other companies raised that are US-focused, digital health, B2B — we're more on the lower end of what they raised. We raised $120 million, right? Just for listeners: competitors raised $500 million and a billion.
Wow.
Yeah. So that's the kind of size of the checks we're talking about, which is completely insane. But then if you look at the size of the problem we're addressing, it's insane too. The dollars spent for pain management in the US alone are like $600 billion a year — $300 billion, whatever you include in your calculation.
Wait, so if I'm understanding correctly, the reason a solution like Kaia is so expensive to build — even if relatively it's not, but it seems quite expensive — is because of go-to-market, rather than the actual computer vision you're making or the clinical trials you're running. That's not the really expensive bit. It's the go-to-market.
Yeah. So the more detailed answer is: you need a couple of geniuses for the computer vision piece. You need time for the clinical studies, which is money. You need to keep existing as a company until you have the clinical trials and the health economics outcomes — because health plans want to know, OK, does it help clinically, but does it also save us money? And then there's the go-to-market sales force you have to invest in. You basically have to figure out the best sales channel, and that iteration takes a while too. So once you have all of that in place and can effectively sell, you need to invest a lot until that moment. All that iteration, all the time you have to wait, and then you get a pilot but you have to wait for the rollout where you actually earn money.
So I think if I'd done it as a first company, I think I would have failed, because it's really hard. Healthcare is really hard and takes a long time. It taught me a lot about patience, and about raising money — because you just have to.
So you've had a number of successful clinical trials and also health economic studies. What I always wonder is — OK, bearing in mind it's a digital health solution, so it's not going to be the same as running a trial on a biologic cancer drug, but still, it sounds kind of expensive. What do you do if the trial comes back and shows that you're not very good? Then aren't you kind of fucked?
You would be, yeah. And the interesting thing, though, is digital health solutions have a retention problem — but if people do it, the outcomes are amazing. So there's no way a clinical trial comes back with bad results. That's the nice thing for us. And that's the weird thing I don't understand until today: if you had the results we have in clinical trials, and no side effects — we don't have any side effects — you would be a blockbuster drug if you were in biotech. But since we're digital health, we're not. The reason is that people still have to do things, and there's not this naturally high price tag associated with a branded drug. But from the impact we have — if you just ignore the word "app" and replace it with "medication" in the clinical trials, and gave that to a pain management physician, they'd be like, that's the best drug I've ever seen.
So it sounds like you should just be hiring these kinds of addiction-marketing type people from, I don't know, Google and Meta — those sorts of people — because you have a great product, but it's about getting people to use it, right? And getting habits formed.
Yeah. So I think we're pretty good at retention right now. Right now, we're more focusing on how we get more of the members' population — the members of the health plans we work with — activated. So it's more like activation right now, instead of retention. And that's enrollment marketing, and being really creative about letting people who had pain issues in the past know about our product, which the health plans know about. So there's a lot of detailed integrations and initiatives we're doing right now. But right now, for us, I don't think it's very hard to be a very profitable company, because we've figured everything out. We know exactly what we have to do to get to a high activation rate with customers. So it's just about helping the customers to do all of that.
I want to ask you about this entrepreneurial gene concept, because your brother is also a very successful entrepreneur. And Scott Alexander wrote this really interesting piece on the "hero license". He was talking about great families in history. You had Charles Darwin, who was related to Francis Galton, who was related to Sir George Darwin. Then you have Niels Bohr — I think his son also won a Nobel Prize, and the other son was an Olympic athlete. The easy thing to say is, oh, it's just genetics, or it's the way you raise the family. But actually, a lot of these people never met. They were family members, but one of them might live in the UK, one in Australia. And the conclusion this guy came to was that it's about the hero license: just seeing someone in your family who looks like you, who is from the same bloodline, and seeing them succeed, makes you think, hey, I can do it as well. Have you got any reflections on that, and maybe how you create a great family?
“There's no shortage of ideas. There's just a shortage of people who love doing things, or who rediscover that they actually loved that when they were young.”
Konstantin
So I think it definitely — when we were growing up, our father is also a founder, and he has a lot of founder friends. When we were growing up, those people were always the happiest people we met. They loved to talk about what they were doing at work. And we were always surprised that most other people really didn't want to talk about their job. If you insisted, they would say, well, it's just to get a paycheck. And I was like, well, it feels like going to a school you really don't like, and that seems terrible. So I think that was one thing.
The other thing is that we did a lot of active things when we were growing up — building things, doing outdoor things, playing instruments. Our parents always encouraged us to do that. And that's what entrepreneurship is about: you just have to do a lot of things. I know a lot of people who are a lot smarter than me, but they're good at analysis and at explaining great strategies. I don't even do research. I just do things, and I learn from one user. If I focus on you as the first user for a product, I'm obsessed about understanding everything in your daily life. And it's just n equals one — but it works, right? Then you just build it, come back to you, show it to you. I just enjoy that process. It's a really happy way of living, because you're building stuff. Almost all children love building. So I think we just get trained to not like that so much anymore, and to value more writing long papers that never get read.
A lot of great researchers have so many ideas out there, and no one does anything with them. So there's no shortage of ideas. There's just a shortage of people who love doing things, or who rediscover that they actually loved that when they were young. The education system messes them up in a way that they get trained to think they don't like it anymore, or that something else gets paid higher — and that's why the best graduates always went there. That's why I wanted to do an internship in M&A, until I saw that it's really not interesting to me, because it's so theoretical and you're not building anything.
So I think the countries that encourage the best graduates to start companies will have a good future. And all the countries where the best graduates still go into service jobs — who do the service jobs want to service, if there are no innovative new companies? That's also one reason I moved to the US: I feel like, especially as a young person, you get an opportunity to do things. I still think in Europe people are held back. I know so many great founders — I invested a lot in early-stage companies — the founders are so good, have amazing products, are great at selling. But they don't get big deals done, because those often don't exist in Europe. A big company wants to buy from another big company in Europe. In the US, if you can show that you save them more money or make them more money, then they give you at least a pilot. And if you perform really well, they expand that pilot. So that's what I like here. But I answered a question you didn't ask.
No, it's great. I have so many more questions, but I think we're up to the end, so I'll stop recording. But that was incredible.
Awesome. Thanks a lot.
I hope you enjoyed that episode. You can find all my links by going to bigpicturemedicine.co.uk. And if you've been enjoying the podcast, then a favor to ask: please consider sending this episode to anyone in your network or friend group you think might enjoy it. It really helps. Thank you, and see you in the next one.