Latest / StartUp Health NOW Podcast / Rethinking the Rules: Julie Yoo of a16z on Infinite Healthcare, Compound Businesses, and the New Investor Playbook

Transcript

Rethinking the Rules: Julie Yoo of a16z on Infinite Healthcare, Compound Businesses, and the New Investor Playbook · StartUp Health NOW Podcast

• 0:00
  [intro jingle] Welcome everyone. My name is Unity Stoakes, co-founder at Startup Health. It's great to be with all of you today.
• 0:12
  We are thrilled to welcome Julie Yu back to Startup Health for another Fireside Chat. Julia is a long time friend of Startup Health and to the whole ecosystem. Some quick background.
• 0:28
  Julia is one of those amazing people in healthcare
• 0:32
  who has been both a founder, a builder in the trenches, and is now also one of the most influential investors shaping the future of the industry.
• 0:46
  Julie is a general partner at a16z on the health and bio side, and previously co-founder and chief product officer at Kairos, which she built and scaled a health tech platform to twenty million patients and over two hundred and twenty-five thousand providers before it was acquired.
• 1:07
  Now, I bring this up because Julie's been through the journey that many of you all are in now. So now at a16z, Julie's leading investments in some of the most talked about AI health companies.
• 1:20
  She's also one of the most active voices in the ecosystem right now from a thought leadership perspective.
• 1:26
  Maybe we can cover in a few minutes your piece on infinite healthcare, which I thought was really important, laying out a vision for how AI can fundamentally restructure care delivery and payment models.
• 1:40
  Uh, but first, welcome Julie. Hello. Thank you so much for joining us. How are you today? I am so well, thanks. Thank you so much for having me again. It's great to see this group, and thank you all for joining.
• 1:50
  I'm really looking forward to this conversation because it really is a seminal moment in the world, but also in healthcare and health innovation.
• 1:59
  We are in the age of super intelligence, where every business, every product, every experience is either being reinvented or needs to be reinvented, and you go to bed and wake up and the world has changed.
• 2:16
  So [chuckles] it's a dynamic time. But maybe just to start, describe the moment from your vantage point within the context of the investors landscape and a16z. What's going on? How do you define the moment today?
• 2:33
  It's incredible from so many different perspectives. You mentioned I used to be a builder.
• 2:37
  I'm a builder again in many ways because I'm, like, once again hands-on-keyboard, just given, like, all the amazing tools that AI now has to offer.
• 2:44
  And so that's actually been, like, very fun to think about all the things.
• 2:47
  And pe-people might have seen my post from a couple weeks ago where my co-founder and I challenged each other to rebuild our entire tech stack from Kairos that we had spent many years building and lots of venture capital dollars building in ninety days, and I joked that I was able to finish it in four days, uh, which is, like, ninety-five percent true.
• 3:04
  And just the amount of kind of surface area and, like, just being generative in and of itself is just, like, a new skill set that I think opens up, like, so much amazing opportunity for entrepreneurs in this day and age that I'm jealous of.
• 3:16
  So that's, like, from a builder perspective.
• 3:18
  From a overall healthcare perspective, I've talked a lot about the fact that I do genuinely believe that healthcare will be the industry that benefits the most from AI because of all the structural challenges that it's had that used to be a liability for us.
• 3:29
  The fact that, that we were the lowest adopters of technology historically. We've had the highest labor-specific challenges.
• 3:36
  We have so much esoteric domain expertise that is necessary to do things responsibly and with rigor in our space.
• 3:44
  Like, all of those things are exactly the criteria that make our industry the most conducive to the application of AI, and we're seeing that play out.
• 3:52
  And I think the slope of the curve of a lot of what we're seeing happening in healthcare specifically is steeper than a lot of other industries which historically have been much more sort of strong beneficiaries of previous tech waves.
• 4:04
  And then venture capital itself, like, health tech venture capital in particular, there was an era during COVID that we all remember, and I think that was probably the last time I spoke with you guys was during the peak of that crazy era when we did see a lot of people come in to investing in health tech because it was part of the zeitgeist and everyone w-was experiencing healthcare full throttle in and out e-every single day of their lives.
• 4:28
  And so it became a popular space to invest for non-healthcare investors.
• 4:33
  I always recall this extremely stark stat that said, I think it was twenty twenty-two, the percentage of deals in health tech that were done by first-time investors in health tech was, like, forty percent, like forty percent of deals were done by noobs to our industry, which ended up actually being a bad thing in that era because as soon as the markets pulled back, everyone went back to their home base and forgot about their k-kind of temporary interest in health tech.
• 4:59
  I think there are signs that this time it's different. We are seeing an influx of capital interest in our space from tech firms and generalist investors.
• 5:07
  But the form factor of the businesses that are now being built in our space, I think are much better set up to be able to take advantage of that influx as opposed to in the previous era.
• 5:16
  I don't think that was so much the case. All that to say, Unity, I could go on and on for all the reasons why it's so exciting right now, but I think certainly from those vantage points, it's been night and day. Yeah.
• 5:28
  Historically, healthcare seemed like it was a laggard in terms of innovation, and as you described, healthcare in this age of AI, new age of AI-One of the biggest opportunities is within healthcare. Absolutely.
• 5:44
  So a few like anecdotes to describe what's happening there. One is just like the funny thing when...
• 5:49
  And I've said this before like in public that when I first joined the firm, I first joined Andreessen in 2019, and our model is that we're a generalist firm, so we have obviously expertise across all sectors of the economy, but we do have specialized funds.
• 6:01
  But when I joined the firm, we were small enough that all the partners sat in one room, and when I would bring my health tech deals to the partnership, I was slightly embarrassed.
• 6:10
  I was like, "Sorry, guys, these ones don't have as great g- growth as the things that you guys are seeing in the broad tech universe. Their gross margins aren't as great. They're just slower.
• 6:18
  They need more capital," right? Like, they were always just... They looked worse on paper than any of their tech counterparts.
• 6:24
  Fast-forward to today, my like consumer partners, my enterprise partners are clamoring to see health tech deals with me, right?
• 6:30
  And it's because on paper, like these companies are growing just as quickly as other companies in other spaces.
• 6:37
  They have unit economics and P&Ls that look far more like tech companies than tech enabled service companies of days past. In health tech, we're seeing a lot more non-health tech native founders.
• 6:49
  That's another thing that I would say has changed quite a bit.
• 6:51
  We've all seen the horror stories of people who come into healthcare with a lot of hubris and really get it wrong to the detriment of the rest of the industry.
• 6:59
  But this time around, we are seeing like incredibly high caliber individuals and companies being built from people who are not native to the space.
• 7:06
  We had a sort of a graph that showed like companies that were founded since 2022 and how rapidly they were able to get to 100 million ARR compared to the prior generation of tech companies.
• 7:17
  And then I went and overlaid health tech companies on that graph, and they're keeping pace with the best and the brightest in the broader tech in- industry.
• 7:24
  So there is definitely something happening that is very different in terms of just the adoption factor in health tech.
• 7:31
  I was with a founder the other night, and he described what would've taken him a year and maybe 50 people, he did with two people in 90 days, so similar kind of construct. The pace is faster. The resources are different.
• 7:47
  How should the ecosystem or how should startups and builders be operating differently today than
• 7:58
  you may have advised companies you were sitting on the board of even a year ago or 18 months ago verse today? Yeah.
• 8:06
  There's some like unintuitive things that we're seeing come out of this new capability set, so to speak, that I'll highlight.
• 8:12
  One is everything that we're talking about with respect to just faster product development cycles, that largely is obviously focused on like the technology pieces of a business which, for healthcare, tends to be probably like the minority of things that you ultimately have to do.
• 8:26
  It does not preclude the need to build a sales force, [laughs] like distribute to the market, to integrate.
• 8:32
  Like my-- the joke that I had in my post, which is real, is like the thing I couldn't do was actually like integrate with Epic, right, across all my customers, which still required a hammer and a chisel t- to get through that wall in an appropriate way.
• 8:43
  The other thing that we're seeing is it used to...
• 8:46
  Like, the conventional wisdom was that you should only do one thing, like one product to one end market, and do that really well before you earn the right to then do, you know, another market or another product or another business model.
• 9:01
  Our best performing companies in our portfolio today are all doing multi-- like, they're compound businesses.
• 9:06
  They are either doing multiple products selling into the same end market or the same product being sold into multiple end markets or a-- the same product with multiple business models, like both B2B2C and direct to consumer a- all at once, and doing it well.
• 9:22
  Like, I would be the first to admit that I was the one pushing my founders to not do that.
• 9:26
  That is like the easiest way to die [laughs] as a digital health company, is to try to push two boulders up two go-to-market mountains or sell into two different markets and not fully appreciate the differences necessary to codify into your operating model to get that right, and where I- they've all proved me wrong.
• 9:41
  And a lot of it is because of this kind of AI leverage that they get on, let's call it, 80% of the stack that you need to build to do things.
• 9:47
  Rather than spend calories on those things, they can now deploy it against these kind of parallelized vectors for bringing their solutions to market and having multiple revenue streams, which I find to be fascinating.
• 9:57
  It seems you wake up the next morning, and if you just have a very narrow path that's been reinvented or a feature inside other models, so that's very interesting. What are the investors looking for?
• 10:09
  What are you looking for that's different today? What do the founders and teams need to show up with- Yeah... that is different to really break through and really stand out? So this is another thing that's changed.
• 10:23
  And every seven to 10 years, there's a new standard for what great looks like at any given stage in company building, and that shift definitely has happened very recently.
• 10:32
  Back then, because just funding was more scarce and there was just fewer companies in our space, it was easier to stand out, right? These days at the Series A, everyone comes in with five million ARR.
• 10:43
  They all have the same product. They're all selling to the same market, and it's really hard to tell purely based on the numbers who is actually the breakout winner, and I think that's great, actually.
• 10:53
  I think it's a great forcing function for us to now focus more on the caliber of the founder, the caliber of the roadmap, the ambition of where this company will be in five years, 'cause that's actually where they all do start to differ.
• 11:05
  But it does come back to how do we assess the exceptionalism of the founder, which a- again, I think is a very good thing.
• 11:11
  How are you thinking about the investor journey and when a company should be ready to come to a16 specifically and come to you? Yes. Our sweet spot is classic venture seed and Series A.
• 11:24
  The way our firm is structured is we have multiple fund families. I personally run our healthcare fund, or the healthcare portion of our fund, alongside our biotech fund, which focuses on seed and Series A.
• 11:35
  And this is actually a point of advice for anyone who's trying to figure out who should I go raise from or who should I pitch to, is really understanding that dynamic of if you're a generalist fund, you will be compared to every other tech company-In the universe when a decision is being made about whether to invest in you.
• 11:50
  So that is very different than going to a dedicated health tech fund where you will be evaluated vis-a-vis only other health tech companies. It's harder to... It's just a higher bar.
• 11:59
  Like, you will literally be compared to, like, OpenAI and SpaceX [laughs] if you pitch to our growth team. Our growth team does have a very high bar. I'd like to bring the first question from the group into the mix here.
• 12:09
  Hi, Julie. Very nice to meet you. So I'm Darlene Molly, the CEO of Next Life Sciences. We're a reproductive health platform. We have two male contraceptive products in clinical trials right now.
• 12:21
  So I just want to get your impression on why this area seems to be neglected in the past. Nice to meet you, Darlene. Thanks for the question. I'll answer just generally versus just specific to your category.
• 12:33
  Marc Andreessen has a saying that the world can only absorb a certain number of ideas in any given period of time, and there is just a law of physics element to which spaces get attention in any given wave, and it really ebbs and flows.
• 12:45
  And so perhaps, Darlene, I actually do think that we are seeing an influx of companies in your space, which I think is wonderful. I think part of it's just consumer zeitgeist, honestly.
• 12:54
  I think that there is a lot of things in the general category of like wellness, longevity, reproductive health that are just popular.
• 13:01
  I think there's been a change in tune about families here in the US just culturally, and so there is a moment. And listen, timing is the hardest thing to get right.
• 13:12
  That's the other thing we always say internally is that there are no bad ideas.
• 13:14
  There is just bad timing, and everything that seems like a bad idea at some point in time that gets made fun of, that gets laughed at, like, at some point in the future, it is likely to actually find its day in the sun.
• 13:26
  Okay, next question, Venkat. I am Venkat Ramamurthy, founder of TrackHealth AI.
• 13:31
  We talked about fragmentation, and the focus of TrackHealth is all the set of activities and tasks that need to happen once a patient gets discharged. So my question is, how do you think about AI commoditizing a new era?
• 13:45
  I agree that a lot of what is broken about healthcare is, like, at the moments of transition of care, and so I think what you're doing is super important.
• 13:52
  And maybe to that point, like, I actually welcome AI commoditizing a bunch of stuff, right? There's too much stuff that's not commoditized in our industry that causes cost bloat and challenges around access.
• 14:03
  Don't we all wish that primary care was commoditized so that we had infinite access?
• 14:08
  Like, that is happening, and I welcome the sort of the forcing function that presents for us to actually focus on, like, higher order problems.
• 14:16
  There will always be last mile issues that still need to be done with human involvement, even if, by the way, healthcare is infinite, like the piece that Udity mentioned that we spoke about, like, in a world that we have infinite healthcare, you will still need physician oversight.
• 14:30
  So I don't think that any of us will find ourselves in a shortage of new things to build against that. You mentioned again infinite healthcare.
• 14:38
  Could you just frame the thesis there and just go a little deeper to what that means, why it's important, and I think it relates to what you were just talking about actually. Yeah, absolutely.
• 14:49
  Basically the premise is in today's healthcare world, as we know, more is bad, right?
• 14:54
  Like, when people talk about growth or higher u-utilization of healthcare, it's generally pitched in a negative light, and we think that's gonna change.
• 15:02
  Like, when-- once you do have infinite supply, quote unquote infinite supply, the commodification of primary care, et cetera, people should be using primary care every single day to manage their lives.
• 15:12
  And in that piece, we actually proposed... It was really a payment model piece. We were saying, "How do you pay for that? What is that worth to us as a society if we are able to implement inf-infinite healthcare?"
• 15:22
  And we proposed four distinct payment models. Part of the point we were trying to make is that there is n-no one monolithic payment model that will cover AI. Lots of people love the piece.
• 15:33
  Lots of people, like, totally hated the piece, and they were, like, still in the more is bad camp.
• 15:37
  And my reaction to that was like, okay, for decades, like, everyone in this industry has been shouting from the rooftops, "We have to solve the access problem. We don't have enough access.
• 15:45
  We have a shortage of capacity, shortage of doctors." And then all of a sudden we come saying, "Hey, guys, we think we can solve access," and then [laughs] all of those same people are now shouting, "No, that's bad.
• 15:55
  We're gonna spend too much." And they're missing the point that the point that we were trying to make in that piece was really about unit cost, right? Yeah.
• 16:01
  The unit cost of care delivery is gonna be a fraction of what it is today. You're reminding me of a talk I gave at South by Southwest, I don't know, six, seven years ago, maybe longer, called The $100 Trillion Market.
• 16:14
  Basically, my thesis was that what today globally was or then was a $10 to $13 trillion market, depending how you counted all healthcare,
• 16:25
  should really be $100 trillion expansion market because there's eight billion people at the time in the world, and they should all be in the system, all getting access, and the cost should come way down so that everyone can actually afford that, but the utilization and the market should expand, and people just freaked out and- Yeah...
• 16:47
  thought I was absolutely crazy.
• 16:49
  And, yeah, and there's legitimate reason to freak out in that there will be some period of time, like, when you squeeze this part of the balloon, like, the other part of the balloon doesn't necessarily shrink immediately, right?
• 16:57
  So it is a real issue of, like, how do we get from here to there, but, like, let's at least agree on what the end state is [laughs] and then work backwards from there as opposed to not be aligned on what that end state needs to look like.
• 17:08
  Next question. Hi, I'm Cole from GlucoSense, and we call ourselves the air traffic control for diabetes population health. I'm wondering what you're seeing in terms of s-startups on the clinical side.
• 17:20
  Like, how are you defining success?
• 17:22
  Are startups on the clinical side being held to the same revenue standards and timeline standards as a general tech startup or an admin portal-side startup Thanks for the question, Cole, and congrats on what you're building.
• 17:35
  Yeah, first of all, digital health is held to a higher standard than human-based care, right?
• 17:39
  And it's even worse now with AI because one thing that AI does is because it is so hyper-efficient, if you just put it into the status quo payment system or workflow systems, it will show all of the warts of that existing system really quickly, right?
• 17:55
  So that scares people, and I think that's part of what's causing the sort of backlash around putting a higher standard of care against some of these things.
• 18:02
  But so the places where we are seeing clinical AI grow at the same clip as some of the hyperscalers in the market, one is health system facing clinical decision support, especially those tools that are supported by the FDA and can actually be used for diagnostic purposes.
• 18:18
  The other arena where we're seeing really high growth on clinical solutions is on the consumer side, and this is actually one of the, another sort of thing that like definitely did not work for long periods of time in the history of digital health, as Unity knows, and now all of a sudden it's back in the zeitgeist.
• 18:33
  The, another joke I put out there was anyone who's been in this industry long enough, I see Esther Dyson here. I was like such a fan girl of you.
• 18:39
  I remember you speaking at like the Health 2.0 conferences back in the day, if you recall those. So Health 2.0 was like where all the consumer health companies tried to do its thing, but none of them worked, right?
• 18:49
  But I compared the categories of companies that pitched at Health 2.0, like 2011 or something, versus the ones that were pitching me last year, and it's exactly the same categories, but now they're working.
• 19:00
  And the confluence of reasons why it's like different today, it's everything from funding availability to payment rails that exist today that did not exist back then, to regulatory rails.
• 19:11
  There's like dozens of reasons why, but some of the fastest growing companies in our portfolio are delivering consumer-facing clinical services. Speaking of Esther, let's bring Esther into the conversation. It's,
• 19:22
  it's a real challenge to me, it seems, in healthcare most especially, but in general, is we're still focused on automating the silos and making them much better, but the, the real challenge is integrating the silos.
• 19:36
  And it's not just, "Oh, I want function health helping you put your blood next to your aura, next to your WHOOP," and that kind of thing. It's more, so I'm running a healthcare system or I'm running some part of it.
• 19:50
  My biggest job is trade-offs. Yeah. It's not how do I run the schedule, but I'm trying to schedule the nurses to do this, but now we've run out of techs to do that thing. How do we integrate the system?
• 20:04
  How do we decide what to prioritize? And unfortunately, but now in the real world of healthcare, everything's in short supply. The workers are burned out. What should I focus on?
• 20:19
  How do I manage all these trade-offs? And that's- Yeah... gonna be the most interesting question, I think. A hundred percent. It's the hardest question, for sure.
• 20:30
  I think there are certain pockets where it's easier and you do start to see some movement on first principles redesign of systems.
• 20:37
  But I do think that for all the ills of what's happening with revenue cycle automation, I'm sure everyone's following all the controversy around the fact that now all of a sudden all the providers are really good at filing claims and the payers are freaking out, and like the math is not math.
• 20:50
  Like, you assume that 5% of all claims in the last 20 years were just not paid simply because providers were not submitting them correctly, and now all of a sudden they're getting submitted correctly, like none of the contracts are valid anymore.
• 21:02
  I am seeing promising signs of payers and providers coming to the table to like literally renegotiate those contracts from a clean sheet.
• 21:10
  Trust, at the end of the day, is like the biggest barrier here, but there are pockets of it that are happening behind closed doors in small pilot type settings, but we are seeing that, which I think is a good sign that for perhaps people are seeing the writing on the wall that like we can't just be these two silos that are AI botting each other.
• 21:27
  We need to actually like design a system that makes sense from first principles.
• 21:31
  So that's like an easy one that like, I think it, it's purely administrative, like not that high stakes, so hopefully we'll see some progress there. I think what you're talking about is far harder.
• 21:40
  I think we have to get a lot of things right for us to thread the needle on those things.
• 21:43
  So how can we actually make these like highly rewarding career paths that will cause people to actually like jump in from other places of our economy that may be losing their jobs with all the AI automation?
• 21:55
  And the related point, which I think is even harder, is we gotta fix like the regulation board certification, like that whole thing, because that is... it's a cartel.
• 22:04
  Like, we just blatantly call that a cartel internally, the fact that we expressly rate limit the number of people who can get qualified for these jobs in many instances, and like that just needs to change.
• 22:15
  So these are really thorny issues, of course, that we can't just wave a magic wand and change, but I think we're starting to see individual states move towards signs of models that might take us into this direction.
• 22:25
  So I do think that it'll happen locally before it happens federally, but I think those are the kinds of things that we'll have to chip away at.
• 22:31
  So there is a little bit of this Band-Aid moment where the system's broken, let's just throw technology on top of it to try to fix, like a temporary fix to just get me paid faster.
• 22:41
  But I actually think there's probably a phase that doesn't involve any technology at all that needs to happen and that we're seeing starting to happen in certain pockets. Thank you, Esther.
• 22:51
  Yeah, so much of this doesn't always come down to just pure technology and AI. It's about relationships and deal structures and the good old-fashioned things of building great businesses. Beth, nice to see you.
• 23:04
  Hi, I'm Beth from LifeBio.
• 23:06
  We have a storytelling app that's AI powered that is engaging people that are older and more frail usually, but I loved hearing that you work with companies and figure out how these things can pivot.
• 23:17
  I just wondered if companies come to you in a seed round where it's already built, it's already in place, but they're looking at sales and marketing as their primary need for the seed round.
• 23:30
  Have you seen that in some of your investments?Oh, we've seen it all. I think there's no specific, like, definition of the stage at which companies are appropriate for a seed.
• 23:41
  We've invested pre-product, pre-business model.
• 23:45
  We've invested in a lot of companies that have a product in market that's already generating some degree of revenue, that has validation, and, like, literally everything in between.
• 23:52
  Warren from Wink Biotherapeutics, could you introduce yourself, please? Hello, Julie. My name is Warren Marcus, and I am a co-founder and CEO of Wink Biotherapeutics.
• 24:02
  So in terms of drug discovery, are you seeing that a lot of companies are using AI now? And is it a negative if you are not [laughs] or have not used it in for- for your discovery? Yeah, certainly.
• 24:15
  It's funny, actually, we were also saying recently internally that, like, last year, every company that pitched us was, like, something AI or AI for something.
• 24:23
  Now it's-- I don't-- uh, not as many founders actually say that because you assume that it's almost the internet or whatever, like digital health is just health, like that kind of thing.
• 24:31
  Like, the way I think about it is, like, the best founders are always opportunistic, and if there is a tool that presents itself in- to the universe that they can use to create a competitive advantage for themselves, then they should use it uniquely and differently and cleverly relative to everyone else in the world.
• 24:46
  This has been absolutely wonderful. Thank you for sharing your time, your perspective- Thank you... your energy, and really pushing the ecosystem forward.
• 24:55
  It was really great to hear everything from the consumer perspective all the way over to discovery and everything in between, so we covered a lot of ground.
• 25:05
  Closing words, anything you'd like to share, maybe last perspective on what's coming next, and leave us with some hope for the future.
• 25:14
  I think this is a moment where it's like in the spirit of what Jack Dorsey did at Block, where you have an opportunity to throw all previous assumptions out the window and really build things in a different way in this era, and no one will fault you for that.
• 25:27
  I could not be more excited for this moment. I'm so jealous of all of you who are building stuff right now. It's, like, the best time to be an entrepreneur in health tech.
• 25:34
  Go forth and conquer, and I hope to be able to cross paths with all of you at some point in time.
• 25:40
  Thank you so much, Julie, and thanks to the entire community of Health Transformers for what you do to, uh, transform the world, make it better. So have a great day, everyone. Thank you for having me.
• 25:51
  [outro music] And finally, an invitation to all of the extraordinary Health Transformers.
• 25:59
  Whether you are a startup, someone building the future of health, or whether you are an investor, funder, or buyer of the solutions out there in the marketplace, this is an invitation to you to join the Startup Health member community.
• 26:14
  Learn more at startuphealth.com. Look forward to hopefully seeing you soon. [outro music]