Amazon/One Medical, Galileo and Healthcare’s Future: In Conversation with Dr. Tom Lee

I recently had a fascinating conversation with Dr. Tom Lee, the founder of One Medical, Galileo and Epocrates on Amazon’s acquisition of One Medical, what he’s building at Galileo and his reflections on medical education and progress in healthtech. Tom is truly one of the most interesting people in healthtech. You can listen to our conversation on Spotify or Apple Podcasts.

Tom founded three successful healthtech companies, starting with Epocrates, a clinical decision support tool used by over one million healthcare providers today. He founded and ran One Medical for over a decade and then founded Galileo, whose approach we discussed in depth. 

Below is a full transcript of our conversation which we recorded in two parts. In the first part, Tom and I discuss the Amazon acquisition. And in the second part, recorded before the acquisition, we hit on everything else. 

Part 1   

Tom, thanks so much for coming back on. There's obviously been some big news since we originally recorded. Congrats on that and really appreciate you coming back to chat a bit more. I guess to start, would just be curious to hear a bit about, what was your reaction when you found out about the Amazon One Medical news?

Great to reconnect and happy to close the loop here since a lot's been going on since we last chatted just a few weeks ago. And maybe one caveat, I'm not involved in the deal. The deal hasn't closed, so anything I say shouldn't be interpreted as a pro or con. But just as a crusty, old former founder of the company, I'm happy to give you my perspective. And apropos to that, I'm sitting there doing my morning grocery shopping, a very luxurious errand buying milk and toilet paper, and my phone blows up and that's how I heard about the deal. There's always been dialogue with a lot of the big heavies on how we might be able to partner. At some level, I'm very long on One Medical, as I've said before, so it's nice to see that it potentially has a well known home, but there's a lot of detail into what that actually means.

I imagine that was an eventful morning grocery run. You only get a few of those in your life.

Yeah, exactly. And at some level, it wasn't a shock, but it wasn't a big deal. For some reason, people were sending me a bunch of congratulations, and I'm like, "This is a journey." And so I didn't view this as a specific milestone, per se. If anything, it signals that there is a lot of interest in working on fixing healthcare by large organizations that are sophisticated in thinking about this in a different way. And One Medical, I think is one of the better assets out there that can do it. As I've mentioned to most people, regardless of the transaction going forward or not, a lot of work remains to be done for all companies involved in healthcare innovation. We're hopeful this results in a positive outcome for healthcare affordability and, let's be clear, continued privacy and security. I think everybody's been raising that as a bit of an issue, but One Medical has always focused on privacy and security and hopefully that will continue. But those are questions to be answered over time.

At a high level, as you mentioned, the interest I think is really an affirmation of the role of physical clinics in tackling healthcare. What are your reflections on the role of physical clinics in really making a dent in healthcare here in the U.S.?

It's an important aspect, for sure, at a minimum because it's how we have been habituated to get care for the last 100 plus years. It is the habit location where we prefer to get our care and we haven't as a healthcare system designed something better. So, obviously, in our prior conversation, we talked about Galileo, but there the intent is to do it more independent of that. But there are some important aspects to a physical relationship in a physical location as part of the ecosystem. Our argument in Galileo is there's plenty of infrastructure already, you’ve just got to repurpose it. But One Medical, the way it's designed is really to think about the experience within that box and then, how do you connect with the team from that box a little bit more thoughtfully?

I think one thing you'd done so well at One Medical was build this brand and experience that consumers love, obviously some of these big tech companies, like Amazon, are famously customer obsessed. What to you might the next level of upping the consumer experience look like at One Medical?

Oh, I think it's pretty endless, and I'd hesitate to plant seeds anywhere in this, but I think there are obvious next gen experiences. You talk to anybody in healthcare and entrepreneurship, they'd probably say the same thing. The harder thing is actually doing it, designing it, effectuating it, and monetizing it in the Byzantine healthcare system we have. We can all fantasize about what modern healthcare should look like and, obviously, we're doing a little bit more of that at Galileo, but the harder stuff is actually getting it to work.

I imagine when you were running One Medical there were times where you thought about or maybe had been approached in the earlier days on the acquisition side. Was it something you'd thought about in the earlier days of the business?

Not really. I mean, again, there's lots of exit ramps for an organization and lots of branches and forks. For me, as a mission focused leader, I don't really get too focused on that. It's like you're driving your vehicle toward the mission and then occasionally you'll see a sign on the left and a sign on the right that says, "Sell here, buy there," whatever. And I generally like to focus on, how do I get to my destination faster? But for some folks, this might be a way to get there. It's always a conversation that happens when you start to build a compelling, differentiated service and so I'm just glad that, at least from the current deal that's on the table, Amazon's known for that. And so, from a mindset perspective, there's probably some similarities.

This is a potential closing of one chapter on the One Medical story. As you think about and reflect on that journey, any thoughts or feelings given the long journey you've had with the company?

Yeah. I think a lot of people are expecting this moment for me and I'm like, "It feels like, if anything, a non-moment." But part of it is, the outcome for me is higher quality, affordable healthcare for all. And so that's the book that I'm writing and One Medical is a phase of that for me as a personal entrepreneur. One Medical's story will continue and that is probably a chapter in that story for One Medical, but there are more chapters to be written in the One Medical story and more chapters to be written in the Galileo story. So, to me, I'm just glad that these are ongoing concerns that are making an impact on healthcare, no different than the work we did at Epocrates even longer ago. And I'm just happy to see that a lot of the pharmacists that I worked with way back then are still contributing to the quality of the content in Epocrates today, And that's over 20 years ago. So, to me, having long-term ongoing concerns that are mission focused and can improve healthcare is gratifying. I’m building a family personally and hopefully a family of organizations that'll continue to have a positive impact.

Part 2:

To kick things off, you've had a really fascinating career path: starting with medical training to founding three different health tech companies. I'd love to hear a bit more about that path in your words. And particularly, I've heard you mention before that you entered business school when you were switching out of practicing medicine full-time with this goal of starting One Medical on Galileo. What motivated that from the experiences you'd had prior to that?

It's kind of the origin story of a lot of my entrepreneurship career. I mean, I was a doc looking to just practice medicine and I loved the profession. I just didn't love any of the environments in which I was practicing from small practice to an academic hospital, to underserved clinics. They all seemed chaotically organized and patients and physicians weren't happy with any of the practice models. And I just assumed that there was a reason why we had this crappy system that was super expensive and nobody could tell me why and I eventually just got irritated and I said, "I don't want to practice in this broken system when nobody can tell me why it's so poorly designed." And so I felt there's a better way to do this, I’ve just got to figure it out. So I rolled the dice and fortunately got into business school and that unlocked this new world called business and technology to better shape the practice environment.

I think when you started One Medical, you really were one of the first people to raise venture funding for these tech enabled services businesses when you raised that round from Benchmark back in 2007, I'm curious what that early fundraise process was like. Obviously it's not a space venture funds had traditionally spent a lot of time in.

It's a great question given the context in which we are today where valuations and capital infusion into this space is pretty extraordinary. I don't want to sound like the guy who went to school through three feet of snow and blizzard conditions and no shoes, but it sure felt like that back then. I mean, I had already worked pretty intensely at Epocrates and we had built a successful company. So, that had helped a little bit, but I knew nobody would've funded a De Novo primary care service business with brick and mortar. So, I bootstrapped most of it and got it to a place of profitability. But even then I knew it would be a tough slog and it was a tough slog. There were some traditional healthcare investors, but I didn't want them on my cap table as our early investor. I really thought that high growth tech would be a better fit. And we must have had many, many, many conversations.

You alluded to it in your answer- I think obviously the world looks very different today in terms of tons of these types of companies raising venture money. How do you think about the types of companies that are and aren't a good fit for the venture funding model on the services side?

I think that is where people try to maybe overplay the One Medical story a bit. That is not a great path for many entrepreneurs in healthcare in services. Services are intensely complicated, people intensive and slower to scale than most people realize, especially when there's a brick and mortar component. You have to be able to stomach high intensity growth thoughtfully with margin and that doesn't necessarily happen with an early stage entrepreneur that's somewhat new to healthcare. One Medical technically was bootstrapped for years before we actually raised venture money. The nice thing about services is it doesn't necessarily need a ton of capital and you can validate a lot more without professional capital and then some businesses warrant venture capital if the opportunity's big and it allows it to move more quickly and you've got the right tech and margin thesis, but many other businesses and services do not.

I think people unfortunately overly assign that as a success path, but there's many great debt models in services that can also be applied, but it really just depends on what people's goals are. And some people like the sex appeal. It's kind of like the NBA, raising from a tech VC, but there's other ways to be successful in impacting services. 

Were those early validation points for you mostly around proving the unit economics within a clinic and the care model?

Yeah. I mean, when you look at healthcare services, it's by definition a low margin business. So to raise from tech VCs, you better have some good thesis about how it's a higher margin business rather than I just sprinkle some technology fairy dust on this and it suddenly becomes this high margin business. I think people underestimate how much work it takes. It's like a restaurant. Everybody wants to own a restaurant or a bar, but good luck at making that a high margin business. So, you have to have the same intensity and discipline, especially when we're talking about services. The tech side's a very different equation and it has a very different value prop to the healthcare industry, but most people when they want to go into healthcare innovation, they usually are thinking about the services side.

I definitely want to ask you about the technology side because I feel like you talk to some people that I guess are pretty cynical of the impact tech can have in healthcare and then other folks that want tech to just do everything. And it seems like in the businesses you built, you've had this nice balance of the involvement of tech. How do you think about the role that technology should and can play and where have you found it effective in One Medical and Galileo? 

Certainly with One Medical, we thought about the front end to an office experience, then the digital extension thereafter and some of the internal workflows and all of those are important aspects of it, but sometimes it feels tech first and sometimes it feels human first and that's great, but it needs to be thoughtfully designed. Galileo's a little bit different. We're more radical on the tech, more radical on the software flows and we have two distinct segments that we serve, tech first and human first. Each thing is deliberately designed with a quote unquote front end. That front end could be a human interface or a tech interface, but you have to think through that journey and that's the universal way that I like to do it, which is design that experience interstitially and then make sure the math supports it.

You mentioned the way you're using tech at Galileo and these two distinct approaches, the tech first and the human first, both living in the same company. This may be a good opportunity just to explain to folks a bit about what you're doing at Galileo, the patients you serve and the basic care model approach.

If you look at my career, each thing is compounding on top of each other. It's always been about how do I improve quality? How do I improve affordability? How do we improve the care experience for both the patient and the provider? And so, with Epocrates, we use point of care decision making. With One Medical, it was how do you make primary care work better? Galileo was really much more normative. It's what should healthcare delivery look like across broader and more complex segments? So, One Medical was just a walk before you run business, which is, "Hey, let's do an urban and commercial primary care concept to see if you can build it, scale it, generate margins to attract capital and really build a decent size business." And fortunately, that's continuing to grow and I'm still very bullish on One Medical's opportunity to replace all the broken office based experiences and even video based experiences.

Galileo's more radical in the sense that we're just trying to blow up all the assumptions and say, "If you are person A with these issues, what would be an optimal care experience independent of how current medicine pays for healthcare and how you typically get it?" One Medical is more working within the system as is. Galileo is more independent.

The digital first ecosystem is designed for the 80%. That's the first touchpoint of care and we can handle much more complexity on the platform, while interstitially intertwining with the brick and mortar. And then we have a high intensity home care model designed around the most complex 20%. And that's when I talk about the human first model. There, if you're an uninsured or Medicaid based schizophrenic patient, you're not going on to the app store looking for health advice. You're just going into the ER. So, we need a high intensity field based team to make sure that we can build trust, engage with you and get you into the right places of care. Same thing with your home bound, demented, really complicated, sick. So, there's a very different care model for each all baked around value and quality of decision making and efficacy of the labor.

It's super interesting to be doing both of those within a company, right? I feel like when we look at the landscape, there's folks that are trying to do the virtual first approach, whether it's an Amazon Care or Babylon. And then others take on risk for really complex patients in the home like the Landmarks of the world and other risk bearing entities. What's it like running both of those within the same company? They're two very different care models with very different uses of technology. 

It's weird. It was very unusual early on. At scale now actually it's fun because both businesses compliment each other. And we built Galileo to partner with payers, ultimately. Obviously, large employers as well and individuals, but the real opportunity here is how do you design value-based arrangements for populations, right? Because when we were at One Medical, it was really hard to design a model that could scale with payers just given the distribution of footprint and the capital intensity and whatever. So, we wanted to design a model that could say, "Hey, just give us any random cohort of patients and we can take care of them in the most effective way where they need to be cared for in a capital and labor efficient model." It's a much more population based approach to care. It's dynamic, complicated and intense, but extremely satisfying. The reason why is because I only have so much time and I don't want to do another company, frankly.

By being able to diversify the types of challenges and capabilities we could build on a common architecture, we knew we could build two muscles at the same time in parallel rather than one muscle and then another muscle. We're building two new muscles on top of our old legacy muscle, which is running office based experiences and combining that into a portfolio of services that can service almost any life efficiently and effectively, wherever they may be. It early on seemed a little bit crazy. Now that we've actually gotten traction and proof points, people are like, "This is pretty cool." It's easy to talk about it now, but back then it seemed a little crazy. 

I'm sure when you go to payers and say, "Hey, rather than carve off this specific cohort, just give us a population and we can take care of them across the board," that really resonates. We had Adam Boehler on a few weeks ago and he was saying the reason that these risk-based models have originated in Medicare to start is that everyone that's old generally has some sort of health condition. And so, you can do a similar care model, but within the Medicaid population, it can really be a barbell where you have folks that are pretty healthy and then folks that are very high cost and it's a different care model. I think it's really interesting what you built at Galileo to be able to address that barbell population.

It's a great example, exactly. Medicaid, it is a very diverse population and relatively under resourced. And so, being able to efficiently address low intensity and high intensity populations with a cohesive care model is exactly part of the inspiration for Galileo. We wanted to go after last mile populations, Medicaid populations, rural populations with a model because we knew if we could deliver not just a care model, but the best care model to these lives, everybody else would benefit.

Is every patient a potential Galileo patient or are there areas or parts of the health system that you're not going to engage in?

Every patient is a theoretical patient of Galileo's. Today, obviously constrained by contracts, insurance types and whatever and geographies, but as we continue to scale, the intent is that any life can be effectively serviced on the platform.

One thing I think that's been interesting in the way that you built Galileo is you've said explicitly that you wanted the company to have more tangible results before you were trumpeting everything you were doing publicly. In many ways that stands in contrast to how a lot of entrepreneurs are building businesses these days. Selling companies is certainly important to potential employees and customers and I'm curious how you think about that strategy and decision to be on the quieter side with Galileo?

I don't know why this is my preference function, but you're right. It does not seem to be a common way that companies are built in general, but certainly within healthcare. I think maybe people falsely assume that a lot of hype cycle generates a lot of customers and/or deals and maybe that's partially true. I just think that at least for us, we come into services with a bit more humility in the sense that it is hard and having done One Medical probably even more so. One Medical seems like a slam dunk from afar and then having built it out and scaled it, you realize how many little things need to happen right.

Because of that and because of the challenge that Galileo is attacking – being able to service any life under value and not just move the needle on quality and documentation, but actually on total cost of care – I think we just recognize that the problem is a lot more complicated and takes a lot more time. We just didn't want to set ourselves up for this big expectation failure. I just don't like to create that stress. If you were to get into the psyche of it, I don't like to create stress for myself. Why would I overcommit to something when I know it's really challenging? I tend to work on the problem, get some more confidence on the problem, build some, commercialize against it, generate some margin, and then start talking about it. So, that's why we're here.

Switching gears a bit, I'm curious, you went to medical school, you built Epocrates which is a tool that's used by a bunch of clinicians and now you're shaping these new care models of the future. What are your thoughts on the future of medical training and medicine and the role of doctors and nurses in the next 5-10 years? 

If I were to step back, I would get depressed frankly on the educational system in general, but certainly the medical educational system and don't get me wrong, I've trained in great systems. I love the faculty and researchers, but it's just like everything. The system is outdated. The licensure and regulatory frameworks are a little outdated. The training is a little outdated and the discipline of preparing clinicians for the future seems way disconnected. I've looked at this many times and I get nervous, separate from we're all looking to hire more diverse clinicians and the medical system doesn't produce diverse clinicians. That alone if you look at trying to address health equity is a challenge. There's many flaws and issues with how we currently train our clinicians. At some level of scale, when I have a little bit more time, I'd love to invest in the educational frameworks.

What would the Tom Lee Medical School look like?

I think the basics and the boot camp stuff is still very important, but I do think that we underestimate the art of medicine. We underestimate the critical skills needed to be a great clinician that are inferred, but not explicitly taught: critical thinking, decision making, human and insights, empathy, and some of it happens here and there in schools, but I think one of the most compelling things would be how what we do changes the outcomes. Meaning the processes in which we work has a lot to do with the questions I ask you. The sequence in who I see for what reason. When we're sitting in an exam room, assuming that everybody is seeing us who could see us, you don't realize that there's a bunch of people that aren't booking an appointment with you because they don't have access and that's an issue. How do we address the people, the errors of omission separate from the people that come into your exam rooms?

It strikes me that as Galileo is taking on risk for broader populations, you really are in a position to do a lot of those things. You have full control of the levers. And one thing you mentioned is maybe in a normal practice, you're not seeing all the folks that you might or should be seeing. A really important part of these risk based models is actually engaging patients and building trust with them. What’ve you learned about doing that? 

It's an ongoing lifelong journey, right? Human trust, human engagement, human behavioral change is not a one size fits all model. It is a complex endeavor when we look at engagement, behavioral change and otherwise. I think every opportunity to interact with a different patient type is an opportunity to learn. And so, we focus more on is the team constantly learning and not being monolithic in its thinking.

Do you have an example of one of those learnings? 

I'll start off with a simple and pedantic one to maybe a little bit more nuanced one. Building out One Medical was hard, but also easier in the sense that we knew our customer and target segment pretty narrowly, right? It was an urban commercial professional as the target persona. The challenge with Galileo is there's hundreds of personas and segments. And so, something as simple as realizing that the preferred communication contact is not email, but phone, right?

The other maybe more nuanced thing that we're seeing is when you have folks that have really significant mental health, behavioral health issues, all rational programs fall out of the wayside, right? And a lot of what we've been doing is getting people who give a crap about people. They figure it out. It's a human to human type of interaction and it's something we can't script and so, some of the work our team is doing in the field is to connect with people that have been pigeonholed. And as a clinician, you understand why. You see a lot of people coming in and out, you pattern match them and you just assume that they have a certain situation going on and there's not enough time and trust built to really understand the trauma, the layers, some of the delusional layers that you have to see through to really connect to the soul on the inside. And our team is fantastic at the commitment to see the human on the other side. And that's something that's been really powerful for our team. And you just can't describe it. It's outstanding work.

You've obviously built a culture that attracts folks like that. Shifting gears a bit, you were one of the early founders in the health tech space. Taking a step back, how do you evaluate what the space has and hasn't accomplished today?

I'm just generally disappointed so far. It's not to point fingers, but I think I would've expected transformation separate from One Medical, which also I think I would've hoped would be broader than where it is. It’s not bad but it's not even at a million lives yet. So, to me, even though One Medical is a great success story, it's still a huge opportunity story at the same time. That's frustrating and then I look at the industry and its ability to move the needle at a material level and that's frustrating. So, I would just say overall, I think somewhat disappointed with pace, but still full of hope and promise that we're going to get to the other side. I just think it's going to take some time. I do think that we're going to get to the other side faster than other countries, maye outside of pure free market countries.

And I do think on the other side is much higher quality, more affordable care for more people that don't come with the traditional rationing trade offs and everything else that you see in other countries. I do think we'll come around to the other side, but I think it's still probably another 10 years. There are really good things that are happening on some services, on some plans, a lot of data improvements, a lot of pipeline type of improvements in tech and biotech stuff. So, there's some really good stuff in and around the tech side, but that services middle is just so hard to unwind and thick as molasses to maneuver through. I think it's going to take some time.

Why do you think the pace has been slower than you might have hoped in the early days of founding One Medical generally in the space?

I mean, One Medical specifically was just building the expertise, the capital, and the talent. It just didn’t exist and more of that exists today, but back then, you're starting from nothing and you're just building stuff from scratch. Now there's a little bit more of a knowledge on how to do these types of things better today. The system has inherent friction points, right? It's consolidated at the purchasing and buying and distribution level for the most part and then it gets highly fractured and localized at the executional level. So because of that, it is a very complicated business flow to navigate. As I've gotten a little bit more sanguine about the market I understand it now and you just take it as reality. Very often when I'm talking with entrepreneurs, so many of them are used to quick hit wins, and if you're really trying to make an impact within the services side, you have to be patient.

If you had a magic wand to fix healthcare in the US, what's one large change you'd make either in policy or some change to the system?

I think I've alluded to this in other conversations. I don't like to get too political because frankly the likelihood for political change and policy change is so low that I think there's a reality that I assume that we're going to deal with the current system as is for quite some time. If I were to say, "Hey, imagine if I could redesign the healthcare system,” I would organize the goods into public and private goods much more explicitly. Public health, preventative health, a safety net should be public goods for both economic and moral reasons. And then a lot of the middle could be private goods.

And specific to the private goods situation, I think the easiest way would be a voucher program to facilitate the consumerization of the care that's already being delivered and already being financed, but just giving a little bit more control to the individual patient. And I think you would start to see more accountability on quality, value and price. So, given that cost is really the predominant concern within the US healthcare system, I think the voucher within the private sector components, not ignoring the public side, would be probably the single change.

So, basically consumers having more first dollar exposure to their care, but not obviously having to take that completely out of pocket, but having some sort of government subsidy for it. Do you think, are we getting closer to a world like that with the rise of these high deductible health plans and HSAs on top of them in the employer world?

I don't view those as quite the same. It is kind of, but in that situation it's really coming out of somebody's wallet, right? And the problem there is everybody's preconditioned that I shouldn't be spending any money in healthcare. So, if it's coming out of your wallet, that's a very different issue.

You were one of the first people to get people around that, right? With One Medical?

Yeah, exactly. And that's very unusual. People don't expect to pay anything out of pocket but if the dollar's already allocated from the insurance premiums and then you're spending out of that, that has a different feel psychically and then people are more likely to consume, but on a value basis, thinking about the total budget and then in theory pocketing the savings there. So, it's a slightly inverse lens on that, which is you start off with the dollars and you spend the dollars wisely and maybe you save a portion of those dollars and then you keep public health, preventive health issues free. So, you don't create a disincentive for the things that people should be doing in the public's best interests and then you keep more of the elective stuff in the purchase pool so that way consumer power starts to drive down prices. 

I think some critics of that might say, "Well, are consumers really best positioned to be making some of these more complex health choices?" Even in employer health plan enrollment, folks get overwhelmed. It's pretty complex, some of these decisions. How would you respond to that?

I'm not as concerned about that. We allow people to buy cars and planes that in theory could kill them if they're poorly built, but there's regulatory frameworks on how cars are built and planes are built and inspected and otherwise. So, hopefully there's a somewhat safety framework around docs and the decisions being made by physicians. Then at some level you're making trade offs on price and quality and reviews and anything else that you would do for any other important purchase. I think we tend to underestimate people's desire to learn when it's their own dollar versus I have to select whatever's covered and whoever referred me to this. 

I'm sure you get to think about a lot of different parts of healthcare and a lot of different solutions on the care model side. Across things that maybe Galileo isn't touching, what are things that most excite you that are happening elsewhere in health tech? 

Yeah. I think there's obviously so many lanes to cover that there's plenty of opportunities still. The areas that I'm excited about in general: obviously life sciences, as a space, has huge trajectory changes for healthcare and outcomes. A little bit more closer to services is data fluidity. You see more organizations focusing on data fluidity, which I think is a really helpful adjunct, same thing with smart devices, maybe some early machine learning on imaging. You’re starting to see some of these early glimmers. 

Well thanks Tom this has been a fascinating conversation - really great to chat.