Some of the biggest challenges for digital health companies today are engagement, navigating the many stakeholders in healthcare, and the approaching but gradual advent of AI and automation, according to a panel of investors who spoke last week at Partners Connected Health Symposium in Boston, Massachusetts. The panel was moderated by Lee Shapiro, a managing partner at 7wire Ventures.
Emily Melton, a partner at DFJ, says that how well a company comprehends the complexity of working in healthcare is an important question to ask about a potential investment.
“I look for people who recognize the ecosystem, who aren’t scared off by the complexity, and have started to work with different partners,” she said. “Sometimes if someone’s had past experience, you can see entrepreneurs who are excited to take these technologies and do something different. But more often than not what we get really excited about is when you see really fresh technologists who are building beautiful products, but are humble enough that they realize they have to seek out and get advice from people who have done that before and have diverse teams that are helping them navigate that ecosystem.”
Equally important, though, is a commitment to proving outcomes. Casper de Clercq, a general partner at Norwest Venture Partners, said that sense of rigor drives his investment decisions.
“I bring a more critical and certainly objective view around, 'show me a randomized control trial, show me the actual impact. Don’t show me a pilot',” he said. “Maybe I have less imagination than others, but I think it’s really important that startups really think about the pathway to get there. Given historical odds, 20 percent of startups will succeed. And you want to be in that 20 percent.”
One of the biggest decisions a startup makes early in its life is whether to sell to patients, payers, providers, pharma companies or some other stakeholder. Dr. Rich Gliklich, an executive in residence at General Catalyst, broke down the pros and cons of some of the different groups, starting with providers.
“[Health systems] are hard to get into but once you’re into them you never go away, so there is that tradeoff,” he said. “Pharma is much quicker because brand teams at pharma have a lot of money to spend around the branding of a new drug. So as long as they can execute around legal barriers — and that can be a really difficult thing — they’re interested. Payers have more focal decision-making, but they are very conservative. They’re thinking about it for years [before they make an investment]. And patients don’t really feel like they should spend more money out of pocket. So each one has its own challenges.”
One of the biggest technology trends impacting digital health is AI and machine learning, and panelists went back and forth on how close AI is to bringing automation to healthcare. de Clercq said that Norwest has a group focused on “semi-automation”.
“Our focus is on what kind of routine things can we take out of healthcare so that physicians can do what they’re best at – which is problem solving rather than routine care,” he said.
Gliklich drew an analogy to high-volume manufacturing.
“Do you start with a fully automated line? No, that’s impossible. You start with hand assembly, you figure out what works, and then you go to semi-automation and then you go to full automation. If you come in with a technology-centric view, it’s really hard to do well. … It will be tech-enabled services for the next 5-10 years until the AI gets so good that we get confused about what’s human and what’s technology.”
Furthermore, Melton suggested, people aren’t really ready to engage with virtual coaches yet, even though the growing interest in services like Alexa and Siri might suggest otherwise. That’s because there’s a big difference between a bot that’s on-call to answer questions and one that’s proactively building a relationship with you to try to change your behavior. Companies conflate the two at their own risk, she said.
More broadly, Melton said, the digital healthcare world hasn’t quite cracked engagement in general in the way that consumer services like Facebook have.
“If you look at what works well in the consumer realm, it’s low-impact easy engagement,” she said. “So Facebook found out people really care about photos. And what they care about the most is photos of themselves. So one of the biggest trajectories if you talk to early Facebook employees was tagging of photos. Because if you got an email saying you were in a photo last night, the first thing you’re doing is logging in…I don’t know what the answer is [for healthcare], but I think you have to figure out what it is that consumers care about around their health that doesn’t require them to do a lot, but builds that trust over time.”