State of digital health investment, part 8: Google Ventures

Ben Robbins, general partner at GV (Google Ventures), sat down with MobiHealthNews to discuss how the multibillion-dollar venture firm invests in digital health companies, AI's place in healthcare and how government fits into the sector. 

MobiHealthNews: How do you decide which companies to invest in? You have a background in psychiatry. Does that influence what captures your attention in a company?

Ben Robbins: The obvious thing is the team. How I think about a team is usually the team can credibly solve some key issues in healthcare. So, from my vantage point, certainly as a clinician previously and now an investor, there are massive unmet needs all over the place in healthcare.  

I've been spending a lot of time as of late in clinical services, and in the services world, those businesses are intense, and you kind of have to walk a tightrope to make sure you're able to build something that's both economically sustainable, but also clinically impactful. And so, the teams have to come from some super credible background. 

It's hard for somebody if their first job is founding a healthcare services company. They're going to take some serious coaching, which we are up to doing on occasion. Usually, the people who are building out healthcare services companies have pretty deep experience in the field, and we are essentially helping them identify and then build out companies that meet some key needs, and they can credibly achieve that.

MHN: To develop a digital health platform that's going to be beneficial, you need coaching a lot of times, right? Even the most experienced people might need coaching.

Robbins: Yes, for sure. Well, that's the thing: All of these companies take a lot of work. Nothing in healthcare is easy. And so, yes, all of these companies take a significant amount of coaching and experts around the table.

MHN: What does GV offer that differs from what other venture capitalists can offer?

Robbins: GV and me, we're full of clinicians. So, we have a lot of people who have sort of been at the front lines and can see the detailed clinical needs, both for providers and patients. I think that's a big piece of it. The other piece is that we're very patient and that we know what it takes to build these healthcare services companies. We're not looking for rapid transactions of any kind. We're really looking to build out impactable companies, and we're willing to do whatever it takes over whatever time period that takes.

MHN: What are some companies that approach you doing right, and what are some doing wrong?

Robbins: For what they are doing right, they have identified some population that is not being well served, of which there are many in the healthcare system, and they're able to clearly articulate what it is they're doing to better serve them, which sounds simple, but that takes a lot of skill to understand. 

For example, where do the contours of this key fit into this lock? And you can unlock a whole bunch of clinical value. You can deliver something that has not been delivered to some population already. And that takes somebody who has their ear to the ground, they've kind of been through the wringer usually, and they understand what it takes to build a healthcare company.  

The thing I see people doing wrong is almost the inverse of that. They have basically identified that there's some need. Usually, it's some kind of access issue, and then they start trying to raise capital. And I love those conversations. I love working with those people, but it takes a level of detail that's pretty deep in order for it to make sense to really launch the company in terms of actually, like, researching spaces. We love doing that. We love working with entrepreneurs, pre-launch, in terms of actually raising venture capital.  

I certainly advise that it is better for everybody when there is a very deep level of detail and understanding of what exactly you're doing and for whom. We're in a funny spot where VC can be thought of as a feather in the cap, and that's a dangerous mindset. The VC is really just a means that's able to let somebody accelerate and build out their idea. But VC is not a very good goal; it's only a tool.

MHN: Can you expand on that? How is VC only a tool?

Robbins: VC is an incredible source of both money and expertise at a very early phase of the company life cycle. So, it's a group of people that have access to the ability to invest and usually have either a network of deep experts or are deep experts themselves and are willing to work at these early stages of a company build, where things are going to be messy, things are likely to fail. It's risk-takers who hopefully have a lot of experience. 

MHN: How does your psychiatry background influence how you shape companies or does it? 

Robbins: Psychiatry is a pretty unique field in a number of ways. One is that the providers tend to behave a little bit differently than the rest of medicine, where psychiatrists and mental health providers, more broadly, tend to be a little bit more free-flowing between more classic healthcare institutions and then working either independently in small groups or in private practice. 

And so, the mindset of mental health providers, I think, is a little bit different than the rest of medicine, and it's really helpful to be able to both have a network of people that I can go to, to make sure I'm tracking trends, but also having been through that myself, having a deep knowledge of the mindset of providers.  

The other piece of it is that I find in venture capital that, historically, many VC batch companies are building services or software for people that resemble VCs. It tends to be either relatively high-end or consumer-oriented. In the field of mental health, a lot of the need is in populations that don't look like VCs, people that have been through trauma and social chaos, and there's addiction mixed in. 

So, I think having that sort of frontline experience gives me a little bit more insight into what real value drivers are, especially in behavioral health, where the patients look different than the population of entrepreneurs and VCs that they usually work with, and the clinicians themselves behave a little bit differently than the rest of healthcare.

MHN: As one of the heads of GV, what have you learned over time? How are digital health companies expanding and flourishing within the market?

Robbins: Honestly, the thing I've learned is that there is this tremendous need for, I think of it as a product mindset, but it's not a traditional sort of software product. It's this idea that you really understand the end user, what their needs are and what the solution is in a fine level of detail. And it sounds basic, but it's really taken me years to internalize this sense that it really matters. 

Who exactly are you serving? If we're talking about rural primary care, that is not an effective level of detail. It really needs to be like, who exactly is the person? Is it the provider? Is it the patient? Is it the insurance company? What exactly are their needs? And I think that it's relatively rare for somebody to get into the level of detail where you really understand. I need you to describe in plain English who is the user of this and what are their needs. What is this exactly helping with? That level of detail is critical for building things that I think of as hugely impactful.  

There are a lot of companies that go out, and they can be relatively successful, but they're not huge successes unless they deeply understand how to unlock value.

MHN: For a VC, you want that level of detail; that makes sense. But for the general public, does that specificity narrow down the ability of the company?

Robbins: This isn't really the criteria for VC as much as if I were going to build a company; this is how I would approach it. So, I'm not thinking about it as an assessment, as much as this is how I recommend building out a company. This is how I would approach it – just really deeply understanding that end user.  

What I find is that I'm much more excited, especially at the early stages, if somebody is really able to just identify, again, like the detailed description of that end user, and the total market may appear to be small, but if there's truly a need and you have a solution for it, I find that usually the total market is much larger than it appears. 

If you're really able to provide something of value to somebody in the healthcare ecosystem, it unlocks a whole bunch of populations, adjacencies, and things that come later. But the first step is not total market, as much as who exactly am I helping, and how am I doing it?  

When starting a healthcare company, I tend not to be super focused on the total market. That can be a little bit deceptive. Often, the markets are hard to see, and, because we're building out the premise of what we do, we're building out services that do not exist. It can be really hard to identify total market, because total market refers to things that do exist and who is being served currently.

MHN: AI has become a big part of healthcare, but some argue that it is hard to determine which companies have long-term, valuable AI and which do not. Do you consider what kind of AI is used within companies pitching to you?

Robbins: The part of AI I'm most interested in is a group using AI to drive fundamental value. If you're really proving that AI can do something, automate some back office functions or some clinical tasks. 

The top of my pyramid of things that I'm most interested in is ... not that they are selling an AI algorithm to somebody else to derive value in some way, it's that they are powering something that is providing value. It's a back office function where they are able to sell to a provider or a payer. If AI really makes things more efficient and reduces cost, they are the ones that are realizing that value. I think that's more interesting.  

When you start getting into selling AI capabilities to somebody else to actually derive that value, I think the jury's out. 

AI is likely to do a whole lot in healthcare. Is it likely to do a whole lot in pure AI software that is selling to somebody else to derive the value? I'm not as sure where that's going to go.  

MHN: What are some examples of how GV invests in or seeks to invest in the payer/provider space?  

Robbins: We're starting to invest in a smaller number of companies with a much more concentrated allocation of both capital and our expert networks, and the experts within GV are super willing to do healthcare services. 

We, of course, love software solutions, but we're quite comfortable doing healthcare services and that is where we are spending a lot of time because what we're seeing is we're at sort of the beginning of the ability to actually deliver high-quality healthcare services at lower costs.  

CMS has done a wonderful job, basically building out infrastructure that starts to blend who is delivering the care and who is taking the financial risk, but we can't assume that CMS is going to actually drive the solutions. They're really just creating the framework. I think we're now just starting to see people who really understand the levers you can pull. 

We're seeing that you really can't expect a doctor to receive their pay differently, and then all of a sudden they're going to adapt their practice in some way that's good for them and their patients, and society at large. What we really need is creative solutions that allow doctors to do what they're good at while leaning into some of these payment mechanisms and capitation, it's really just various flavors of capitation, and basically help clinicians deliver care in a way that maintains or improves quality and decreases cost.

MHN: It's hard for the government to keep up with the pace of technological advancement. Is it necessary for the public and private sectors to work together?

Robbins: I think the government has a role to play with technology insofar as they are making sure things are safe and of reasonable quality, but I think that the key role the government plays, especially in healthcare, is building flexibility in reimbursement. 

It's a mistake to assume that the government is going to directly drive innovation. I don't think the government wants that, and I don't think entrepreneurs want that, but I do think there is a common sense that we're waiting for CMS to develop some program that is going to drive innovation. I don't think that's what CMS wants, and I don't think that's what we should want either. 

What we should want is for CMS to build flexibility in reimbursement, and then we rely on industry, on academic medicine, on payers to actually build solutions that are unlocked by this flexibility in reimbursement.