Running Through Walls · 2026-06-30 · 14 min
Key moments - from our scoring
Substance score
52 / 100
Five dimensions, 20 points each
The hosts analyze survey responses from healthcare industry participants on major 2026 trends. On AI, 73% report growing trust in AI-assisted medical care, with personal experiences like Siobhan's health coaching via Claude illustrating practical adoption. The GLP-1 discussion reveals a bifurcated market: oral formulations removing injection barriers, but payor coverage battles driving direct-to-consumer adoption with prices expected to drop to $50-75 monthly within two years - mirroring the economics of statins and ED drugs. Peptides emerge as a speculative bet despite lacking FDA approval and uncertain efficacy. The M&A outlook remains bleak, with no meaningful IPO market expected until 2028, though 54% optimistically predict revenue cycle management exits. A Nature paper demonstrating that general LLMs (ChatGPT, Claude, Gemini) outperform specialized tools like OpenEvidence signals that business durability in AI-enabled healthcare is now measured in months, not years - a fundamental shift forcing portfolio companies to continuously pivot rather than build sequentially.
Around $150 per month provides the threshold where payors remove coverage restrictions, with further compression expected to $50-75 monthly within two years as generic competition and rebates intensify.
Yes, a recent Nature paper showed that general frontier LLMs including Gemini, ChatGPT, and Claude significantly outperformed healthcare-specific tools like OpenEvidence on clinical intelligence metrics, even using six-month-old model versions.
Survey respondents predict 2028 or later, with the majority essentially expressing uncertainty rather than near-term optimism for healthcare IPO activity.
Oral dosing eliminating injection barriers, cash pay pricing becoming competitive with net payor prices, and strong consumer demand are overcoming payor reimbursement battles and coverage limitations.
Business durability has compressed to less than 12 months as frontier models replicate functionality, forcing rapid pivoting and deeper verticalization or risk obsolescence and valuation collapse.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode contains a handful of genuinely substantive moments - the GLP-1 market structure analogy, the LLM-vs-OpenEvidence Nature paper, and the AI durability problem - but roughly half the runtime is consumed by personal anecdotes, insider jokes, and banter that adds nothing actionable for an operator.
my bet is is that the profit margin in the GLP-1 market gets squeezed pretty hard on both ends of the spectrum. Both the generic side - you know, to pennies on the dollar - and the rebates on the high-end side
you used to have your first-horizon business last five to seven years and during that time, you can figure out what you're going to then build on top of that. And now you have less than a year
The GLP-1 market comparison to the statin/ED convergence is a genuinely useful and non-obvious framing, and the point about frontier models collapsing the value of specialized healthcare AI products is sharp and timely; most other takes are competent but unsurprising VC commentary.
the GLP-1 market is going to end up looking as though the erectile dysfunction and cholesterol-lowering markets had a baby
we are seeing time and again that the hot company from a year ago is the product roadmap for the large frontier models this year
Bryan Roberts and Bob Kocher are legitimate, senior healthcare VCs at Venrock with real portfolio exposure to the companies and dynamics they discuss, giving their predictions genuine grounding; however, this is an internal team roundtable rather than an interview with an operator or founder who has executed at scale.
many of our portfolio companies have already discovered that because they're delivering care and AI is enabling it
The most shocking M&A transaction that I've ever heard of where they didn't recut the price when the revenues were going to fall by 15%
The episode anchors discussion in their proprietary survey data with named percentages, a specific price threshold prediction ($50 - 75/month in two years), the Nature paper as evidence, and a named deal (Aetna/Oak Street); the predictions are directional rather than rigorously evidenced, but the specificity level is above average for a short-form VC pod.
54% of people believe that revenue cycle management and AI-powered administrative tools will have promising exits in the next 12 months
all of the top three LLMs - so, Gemini, ChatGPT, and Claude - performed better than OpenEvidence, significantly more so on clinical intelligence metrics
The format is a casual internal team chat with minimal probing; Bryan occasionally pushes back ('How do you know what before was?') and asks a useful follow-up on payor coverage restrictions, but most exchanges devolve into affirmation loops or jokes rather than productive challenge or depth-seeking.
How do you know what before was? You don't know what before was. It wasn't in the question
Even the coverage restrictions, right? Like what's been going on in the first half of this year. You think those dissipate?
Computed from the transcript - who did the talking, and the words that came up most.
The results of Venrock's 10th annual Healthcare Prognosis are in. This year's survey covers some of the biggest questions in healthcare, from AI in medical care and the future of GLP-1s to consumer peptides, IPOs, M&A, and what's ahead for healthcare policy. In this episode, Bryan Roberts, Bob Kocher, and Siobhan Nolan Mangini sit down to discuss the most interesting findings, share their perspectives on what surprised them, and offer a pulse check on where the industry may be headed next.
Transcribed and scored by The B2B Podcast Index.
Bryan Roberts: Hey, everybody. Welcome back to Running Through Walls. Bob, Siobhan, and I are here today to talk about the results of the 10th annual Healthcare Prognosis. It�s always exciting.
Bob, Siobhan, welcome. Thank you.Bob Kocher: It�s fun to be here. It�s always exciting because it�s fun to write the questions.
And then, sometimes, the answers are also interesting. Let�s jump in.Siobhan Nolan Mangini: All right, let�s jump in. So, let�s start with the topic on everyone�s mind, AI.
So, 73% of our friends have said their trust in AI assisting their medical care has grown over the past year. This is very true for me. Bob Kocher: And Siobhan, you�ve fallen in love with your AI.Siobhan Nolan Mangini: You both have heard about how much I love my coach.
I think it will change not just my own way that I interact with my healthcare, but with many people�s. Maybe just a story. So, I�ve put all sorts of things into my health coach, including all my lab results. I got Function, which is a way that you can get hundreds of lab results.
Super confusing though, actually, when you�re trying to deal with it. And I put it all in. And basically, the TL;DR was don�t worry about anything. So, it was very helpful.
Bryan Roberts: Oh, that sounds super useful. Siobhan Nolan Mangini: Super useful.Bryan Roberts: Yeah, and I can see why you decided to spend the money and do that. That�s great.
Bob Kocher: What I love is that you �Bryan Roberts: Good job. That�s like I got the one percent risk of CV disease and one percent risk of Alzheimer�s.Siobhan Nolan Mangini: Exactly.Bryan Roberts: I�m so glad I got my genotyping done.
Siobhan Nolan Mangini: Put it into Claude.Bob Kocher: What�s great is that I�ve been telling �Siobhan Nolan Mangini: Don�t worry about it.Bob Kocher: I�ve been telling you, Sioban, that you�re completely healthy for a long time and you don�t believe me. But the AI tells you �Bryan Roberts: How much did you pay Bob?
Siobhan Nolan Mangini: But my coach. But my �Bryan Roberts: How much did you pay Bob for all the lab tests? Nothing. Bob, you got screwed.
Bob Kocher: I know. I should have been an AI coach.Bryan Roberts: So, whenever we ask questions and we get like 18% answers across five different things, I take away nobody�s got any clue. Nobody�s got any clue what�s going to be next, what�s going to hurt it, etcetera.
Which makes total sense given the pace at which AI capabilities are growing. And the pace at which user norms �Siobhan Nolan Mangini: Are changing.Bryan Roberts: � right, are changing. Right?
So, we all were like a year ago, �Oh, no, we�d never do X.� And we�re totally doing X every day right now. The same thing will happen in 12 months. It�s fascinating velocity and uncertainty, in my mind.
Bob Kocher: All right. You said 18% across five things. Sort of true, but not really. Bryan Roberts: Oh.
Bob Kocher: If you add up the top two �Bryan Roberts: Go ahead.Bob Kocher: Seventy-three percent of people basically say they�re comfortable with AI. Some want a doctor to check it. Others are like, �just let it do stuff.
�Siobhan Nolan Mangini: Yeah, clinical AI.Bob Kocher: And I think that�s a pretty exciting change �Siobhan Nolan Mangini: Yeah, clinical AI, for sure.Bob Kocher: � compared to before. Bryan Roberts: How do you know what before was?
You don�t know what before was. It wasn�t in the question.Bob Kocher: Should have asked that question last year; we could have compared.Bryan Roberts: Let�s move on.
Bob Kocher: It�s all going to go great.Bryan Roberts: Okay. I want to do GLP-1s and peptides.Bob Kocher: Which ones are you taking?
Siobhan Nolan Mangini: Great. Let�s do it.Bryan Roberts: No, I�m not taking any of them.Bob Kocher: Okay.
Siobhan Nolan Mangini: I�ve been asked. Bryan Roberts: To take them?Siobhan Nolan Mangini: No, whether or not I am taking GLP-1s.Bryan Roberts: Are you still speaking to the person that asked you�whether you�re taking �Siobhan Nolan Mangini: Oh, it was a random person.
It was literally someone who met me for the first time.Bryan Roberts: Oh, that�s a nice opening.Siobhan Nolan Mangini: Yeah. Yeah.
Bryan Roberts: That�s a great first date question.Siobhan Nolan Mangini: So, I was like, �Wow.�Bryan Roberts: Okay. Good.
No, no. So, GLP-1s� Go ahead. You can do the results and then I�ll give you my take.Siobhan Nolan Mangini: This is one where, I think, people are seeing a lot of heterogeneity in terms of what the outcomes will be.
So, 31% of people think that oral dosing removes the last major barrier now that we�ve got orals that are out there. But 29% think that there will be a slower adoption because of payor coverage battles. I actually see these two things coming together. So, a lot of people want GLP-1s.
There�s going to be maybe some friction on the payor side. And so, we�re seeing a ton of cash pay, direct-to-consumer adoption of GLP-1s. And it says that 21% of you think that there�s going to be strong demand, but challenging drop-out rates just because of adverse events.Bob Kocher: And nausea, yeah.
Siobhan Nolan Mangini: Yeah. And so, we see that. But the reality is prices are dropping really fast and now we have an oral so you don�t have to inject yourself. And so, I mean, maybe what are the thoughts.
Bryan Roberts: Look, I think everybody�s going to want them.Siobhan Nolan Mangini: The pill.Bryan Roberts: Having a pill bail you out of all of your fried twinkies. 2.)
There�s lots and lots of competition. Right? And that competition�s increasing. And you�re about to have the first generic start to come out.
So, I think the GLP-1 market is going to end up looking as though the erectile dysfunction and cholesterol-lowering markets had a baby. Okay?Siobhan Nolan Mangini: Okay.Bryan Roberts: So, you�re going to have � first run is generics for everything.
Like statins�in the market. And then there�s going to be a set of folks who need the brandeds because they have better efficacy and stuff like that. But there�s going to be multiple of those. So, that�s going to look like the old ED market where the rebates are like 60% or more.
Right? And so, they�ve got high list prices, but the net prices � which goes to your cash pay thing � right? Net prices start approximating cash pay. And so, my bet is is that the profit margin in the GLP-1 market gets squeezed pretty hard�on both ends of the spectrum.
Both the generic side � you know, to pennies on the dollar � and the rebates on the high-end side.Bob Kocher: But I think they�re going to be in the water. So, I think these are medicines like statins that works super well, but better than statins and they make you prettier and skinnier. And people don�t like being hungry.
And these take away the desire to have snacks. And so, I think that you�re going to have people demand them and they�re going to get covered because they�re cheap enough. And �Bryan Roberts: Even the coverage restrictions, right? Like what�s been going on in the first half of this year.
You think those dissipate?Bob Kocher: I do. And I think that actually the clinical ROI is going to be quite good. Because weight loss makes so many things better.
It makes heart disease better, liver disease better, cancer risk fall. And so, I think that payors are going to be like, okay there�s enough clinical reasons to cover it and the prices are low enough either because of the cash pay prices �Bryan Roberts: The net prices.Siobhan Nolan Mangini: What price do you think the net price needs to be for coverage to�?Bob Kocher: I think around 150 bucks.
Siobhan Nolan Mangini: So, where we are now.Bob Kocher: That�s sort of low enough and so, we�re kind of there. Anything that�s below 100 �Bryan Roberts: I think it goes lower. I think it goes lower than 150 bucks, yeah.
Siobhan Nolan Mangini: I do, too.Bob Kocher: I think that�s the price at which � sorry, the question � for where payors take the seatbelts off. Bryan Roberts: I think you get to 50 to 75 bucks a month.Siobhan Nolan Mangini: In the next year?
Bryan Roberts: Two years.Siobhan Nolan Mangini: Two years. Based off of generic timing?Bob Kocher: Yeah.
Bryan Roberts: Generic timing, competition, etcetera, etcetera�is my bet.Siobhan Nolan Mangini: Okay.Bob Kocher: Bryan, what�s the deal with peptides?Bryan Roberts: Peptides.
Okay. So, No. 1, oh, my God. Let�s just be clear.
20% of you thought that a peptide therapy would get covered by a commercial insurer by the end of 2026.Bob Kocher: Those are the 20% that have never talked to an insurance company.Bryan Roberts: What on earth?Bob Kocher: They don�t cover non-FDA approved � Bryan Roberts: Are we going to be on Mars in 2027?
Like, it�s about the same likelihood. Bob Kocher: Elon�s [inaudible] so, maybe.Bryan Roberts: So, that was the No. 1 thing that surprised the bejeezus out of me.
A bunch of you � a third of you, nearly � believe that peptides will be hot and on their way to the mainstream. It�s surprising to me. Bob Kocher: It�s surprising to me too because a lot of people are, you know, scared of needles and these are injectable drugs. The second thing is, they�re drugs, but they�re not FDA approved.
Bryan Roberts: You could take them orally. And then they just broke down in the stomach. And they do exactly what they do when you inject them, too. Nothing.
Siobhan Nolan Mangini: Exactly.Bob Kocher: Placebo. Although maybe you have to hope for nothing instead of something because you�re not sure what they�re going to do. I think it�s shocking to me that people are buying these things that make claims that they�re going to make you stronger, skinnier, smarter, or have better skin and hair, take away your pains, when nobody really knows quite what�s in them and how they work and if they�re safe.
So, I�m surprised that the HHS, I mean, is promoting them as much as they are given the risks in the safety and manufacturing. And the expense of these things; they�re not cheap.Bryan Roberts: All right. Let�s move on to IPOs.
Bob Kocher: Yeah. SpaceX. Big IPO. Oh, you mean healthcare IPOs, Bryan?
Bryan Roberts: No, there are none of those.Siobhan Nolan Mangini: None of those.Bob Kocher: I know. Bryan Roberts: We don�t do those.
Bob Kocher: No, but where are all the acquirers?Bryan Roberts: So, and it�s true. Half of respondents believe that the IPO market would come back in 2028 or later. That�s basically saying I don�t know when it�s going to come back and I�m going to say a time that you guys will forget what time I said when you next ask me.
Bob Kocher: That was the last option on the answer actually. I didn�t have 2040.Bryan Roberts: Okay.Bob Kocher: But that made me sad.
Bryan Roberts: But yet, a whole bunch of people answered the next question, which was, �In the next 12 months, which health tech subsectors will have the most promising exits?� Those answers feel incongruous to me. Bob Kocher: Yeah, it�s because we�re going to be exiting, too. Siobhan Nolan Mangini: Yeah.
Bryan Roberts: Exactly. Right? There�s no IPO market until 2028 or later.Bob Kocher: And there�s basically no M&A market.
Bryan Roberts: But 54% of people believe that revenue cycle management and AI-powered administrative tools will have promising exits in the next 12 months.Bob Kocher: I guess they�re all hoping to meet Matt Holt. Siobhan Nolan Mangini: Yeah, I was going to say, I think there�s a hope.Bryan Roberts: Oh, that�s right.
Let�s call Matt.Siobhan Nolan Mangini: Yeah, exactly.Bob Kocher: Operators are standing by.Bryan Roberts: You think we�d get in trouble if we gave out Matt�s cell phone on the podcast?
Probably.Siobhan Nolan Mangini: He might be kind of mad.Bryan Roberts: We won�t do that.Bob Kocher: Okay.
Bryan Roberts: There you go, Matt. That shows you how much I love you. Bob Kocher: But he might be the only exit available, so you better go find him.Bryan Roberts: In news that nobody is surprised about, nearly half of you believe that Aetna will divest Oak Street.
A good three years after it was bought, right on the cusp�Oh, after. After the cusp �Bob Kocher: After.Bryan Roberts: � of V28.Siobhan Nolan Mangini: Yeah.
Bob Kocher: Yeah. The most shocking M&A transaction that I�ve ever heard of where they didn�t recut the price when the revenues were going to fall by 15%.Bryan Roberts: Yeah. Exactly.
Bryan Roberts: So, you guys all think that that�s going to get ditched�Bob Kocher: The big news, I thought, was the Nature paper showing that the large LLMs performed even better than the healthcare-specific AI tools like the OpenEvidence. Now, many of our portfolio companies have already discovered that because they�re delivering care and AI is enabling it. And they�ve actually shifted to currently the new model of OpenAI is the one that many are using � from Claude. But they found that the platform LLMs are awesome.
But the Nature paper was really jarring in that it showed like oh, my God. Even the old LLMs were better than OpenEvidence and UpToDate. I guess, Siobhan, what �Siobhan Nolan Mangini: I thought this is super interesting. And, actually, we have a question where 47% of you thought that OpenEvidence would be disrupted by OpenAI or Anthropic.
This is a question we just put out less than a month ago.Bryan Roberts: Damn, we should have gotten the Prognosis out sooner.Siobhan Nolan Mangini: I know. I know.
But then, this Nature paper hit and all of the top three LLMs � so, Gemini, ChatGPT, and Claude � performed better than OpenEvidence, significantly more so on clinical intelligence metrics. And I think, to your point, they were old models. So, like six-months-plus-old stale models still perform better. And I think this is just demonstrating how fast � something you mentioned earlier, Bryan � is just how fast things are moving and what is actually durable right now.
Because the frontier models can move.Bryan Roberts: Right. The durability question�s a tough one. Right?
Siobhan Nolan Mangini: Yeah.Bryan Roberts: It is absolutely � we are seeing time and again that the hot company from a year ago is the product roadmap for the large frontier models this year.Siobhan Nolan Mangini: Yup. We saw this with the scribes.
Bryan Roberts: Yeah.Bob Kocher: Exactly.Siobhan Nolan Mangini: Right.Bryan Roberts: Right, so, it�s really � it�s like how you infuse AI into a business that will be durable, such that for both management teams and investors, there�s persistence of value creation is one of the � from my perspective � one of the big questions for us and the ecosystem over the course of the next couple of years.
Bob Kocher: You used to have your first-horizon business last five to seven years and during that time, you can figure out what you�re going to then build on top of that.Bryan Roberts: Really do. Yeah. Bob Kocher: And now you have less than a year �Siobhan Nolan Mangini: Yeah, less than 12 months, maybe.
Bob Kocher: � and your price collapses, nobody will buy your first business anymore, and oh, my God. And so, it�s going to be a tough one. And these companies that are also highly valued are going to have a tough time figuring out what to do.Bryan Roberts: Yeah, the rapidity of requirement to pivot and add on is a fascinating one.
Siobhan Nolan Mangini: And get deeper and deeper and deeper. So, it will be interesting �Bob Kocher: But it is exciting that the core LLMs can do really high-end clinical care because it means that there will be a lot more, actually, innovation, a lot more access to it. And the other thing about AI is that they�re more evidence based than people. And so, more doctors use it to augment their care.
That�s going to be good for patients.Bryan Roberts: Certainly, it�s more up to date than the 20-year alumni of medical school. Bob Kocher: But we like going to our reunions.Bryan Roberts: All right.
Anything else, guys?Bob Kocher: Well, midterm elections. People predicted that the Democrats took the House, Republicans will hold the Senate. And my prediction is that that�s probably true and that healthcare won�t be a big deal in the midterm elections.
It�s going to be a 2028 topic because the H.R.1 Medicaid cuts happen after midterm elections. All of the squeezing on rural hospitals happen after the election.
Bryan Roberts: Totally agree. This is energy prices and economy. Siobhan Nolan Mangini: Yeah.Bob Kocher: Yup.
Bryan Roberts: Not healthcare.Bob Kocher: And healthcare is going to be a problem in 2028 because the inflation in healthcare�s going to come back to being a really big problem.Siobhan Nolan Mangini: Absolutely.Bob Kocher: I appreciate everybody filling out the survey.
It�s fun to write it, it�s fun to read the results. And I hope you all find it interesting. Siobhan Nolan Mangini: It was fun to dive in the results and get a pulse check on the industry. See you again next year.
And thanks for listening.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.
#291 Why Most AI Projects Fail to Deliver ROI, Sinohe Terrero, CFO and COO, EnvoyGrowCFO Show · on Claude91 / 100
Why a $1.2B exit felt like his biggest failure, and the customer-obsession thesis behind AgencyThe GTMnow Podcast · on Claude86 / 100
Is Your Business Invisible to AI Search? (And How to Fix It) ft. Ray YoungRevenue Science · on Claude85 / 100
Episode 018: Season 2, the $75 Consult and the Frankenstein StackAI Tools for Practicing Lawyers · on Claude84 / 100
Agentic Engineering for Testers: How to Automate Your Way to the Top with Amit RawatTestGuild Automation Podcast · on Claude82 / 100
PwC's Chief People Officer on Training 80,000 People for the AI Era With Human Skills at the CenterFuture Ready Leadership With Jacob Morgan · on Claude81 / 100Arca's AI-First Vision for Human-First Wealth ManagementBuilding a Differentiated Obesity BiotechThe Long Game of Building Breakthrough BiotechSmithRx's Mission to Break the Drug Pricing Black BoxPursuit over Podium: Rethinking Excellence with Brad Stulberg