Digital Health is Dead. Long Live Digital Health.
When everything is up and to the right and everyone gets a participation trophy just for showing up, it’s easy to be directionally correct and delude yourself into believing that you’re specifically brilliant. But when things get hard and the winners get fewer, a new danger emerges. What if you’ve been fishing in the right pond but with the wrong bait the whole time?
That brings us to the red herring and broken model of venture-backed digital care delivery.
Over the past decade, there’s been a consensus playbook for how to grow and scale a digital healthcare delivery company:
Launch a cash pay business with a narrow aperture of services that are high (“soft-dollar”) ROI for a population — think convenience, user experience, access to care, etc.
Continue growing and achieve sustainable CAC at scale. This is only possible with 1. Some kind of moat, 2. Going in-network with payers and/or 3. selling direct to employers.
Expand into new categories/treatment modalities to grow ARPU and TAM. Now that you have the contracts, the brand, and the tech, you’ll want to leverage those investments against the broadest universe of serviceable revenue.
Once payers fully appreciate the value that you drive, switch from fee for service to some kind of direct contracting that properly rewards you for your special sauce and measurably improved patient outcomes.
The problem is, it’s not clear that this playbook actually works. At all. Sure, lots of these companies were richly valued, with investors clearly expecting them to continue growing rapidly and eventually become profitable. But none of them have actually done it. Livongo started out with employers. Hims went public without ever contracting with payers. Plenty others skip straight to Step 2. Nevertheless this was the broadly accepted wisdom for how to build a digital health business and take it from startup to public company.
But simply putting a doctor’s office on the internet doesn’t fundamentally change its financial profile while adding millions of S&M and R&D makes it much worse.
We believe that irrespective of how something trades for a given quarter or year, quality businesses that can deliver equity efficient growth at scale will be rewarded over time. Everything else and every other kind of business is speculative. So putting aside valuation (how other people/the market views a business) and instead looking at value (first principles analysis) it’s not clear that the conventional playbook/thesis holds water.
Beyond being unprofitable, none of the direct comps even appear to be trending in the right direction.¹ Across the board, the unprofitable ones also have unattractive incremental margins: more revenue widens losses insteading of shrinking them.² Even Progyny, which comes the closest to provisioning care of any of the profitable examples, is not becoming meaningfully more profitable as it grows. There are few or no network effects or economies of scale in providing healthcare.³
But if there’s no copy-cat model, are there other indirect precedents with similar attributes that we can build a case around? In short, no. The consumer “tech enabled” businesses that have really succeeded all pretty much replace a service provider with a self-service tool (do your own taxes, incorporate your own business, book your own hotel, find your own apartment, etc.). Conversely, there are no examples of companies succeeding with models that “supercharge a service provider” — at least not profitably and in the public markets. Why should we believe things will play out differently in healthcare?
In the cold light of day and without the benefit of surging capital markets, it’s clear that investors/founders/the market have been directionally correct (the shift to virtual care and digital tooling in healthcare is powerful and obvious) but specifically wrong (the consensus models and playbooks for provisioning care are not fundamentally sound).
Whether or not it’s backable at the care delivery level (and it seems increasingly clear it is not), if digital care delivery is the arc of the future what IS backable? Where are the equity efficient models and how should we bet on the future of digital health?
We have a few ideas:
Software primitives for virtual care
Despite the challenges of rolling it in-house, software and telemedicine can make a huge difference in the cost and quality of care. Digital therapeutics, bringing providers to care deserts, and improving clinical workflows and healthcare operations all matter and are worthy endeavors. There are basic functions, tools, and products that every digital healthcare provider will need to serve patients. Outsourcing those components to software companies can be a win-win: the providers can focus on care and shed redundant R&D budgets/headcounts while the software providers can be just that. Everyone gets more equity efficient business models. Just like how every fintech company doesn’t build its own fullstack payments, KYC, and account funding infrastructure, digital health providers don’t need to (and shouldn’t) roll their own patient onboarding, patient communications, outcome tracking, care-team coordination, payments and billing, etc. Companies like Capable, Formsort, Fold, and Source are great examples of this. These companies will also help power a new generation of digital-first private practices totally separate from investor owned healthcare.
On the other hand, purpose-built EMRs or EMR developer tools are a tempting area but run the risk of turning into consulting businesses or outsourced R&D that spends too much time/money on customization rather than selling truly zero-marginal-cost software products. Perhaps a sufficiently developer-focused option can get around this risk but could still require its clients to have large engineering teams to implement/run it.
Provider networks
Similarly, recruiting and retaining providers is often a massive drain on resources for digital health companies who are simultaneously trying to build software and provide care. So just as there are opportunities for companies to specialize in software, there are massive opportunities for companies to specialize entirely in recruiting, retaining, and managing provider workforces, especially for high-demand specialities. Progyny is a great model to emulate not only because of its success selling a high-priority/high impact benefit to employers but also because of its success aggregating a constrained group of sought-after providers (REIs). Progyny can charge a premium for access to those providers and knows that its customers will gladly pay it (not that they have a choice). Alma, Headway, and Sondermind are doing something similar with mental health providers. Workforce management and staffing platforms like Wheel (Slow port co), SteadyMD, and Openloop work more like provider marketplaces than virtual clinics. In contrast with someone like Teladoc, they can spend a lot less to grow because they’re making an embedded one to many sale: 1) their clients are doing the selling for them and 2) they also have more motivated buyers by selling to platforms for whom staffing is a fundamental input rather than benefits managers for whom it is an additional cost.
Marketplaces and Self Service Tools
If there aren’t great precedents to get us excited about delivering “tech-enabled” services, then we should look to back parallels for the models that do work in other categories. Namely, self service tools and marketplaces. Book your own appointments (Zocdoc) works but there are probably good opportunities to either un-bundle it or compete directly. The same goes for things like finding clinical trials, submitting insurance claims, ordering DME, etc. Software and software-led (ie not “managed”) marketplaces work. Why not in healthcare?
But if you’re going to build a healthcare services business, focus on solving acute problems for payers/employers via proprietary or built-in distribution. Making something consumers want and delivering it via a great product has to be table stakes, not the end point. You have to create and capture value; the only way to do that is through solving high priority issues for payers, ideally and eventually through direct contracting or value based care. Remember, the ultimate buyer for healthcare isn’t the CHRO, it’s the CFO; you have to prove bottom-line value to payers and employers. Startups just can’t operate on 12–18 month feedback loops so you also need a way to escape the payer-sales-cycle vortex of death. Progyny and Livongo were both quasi-incubated and launched with employer relationships to avoid ever being DTC or dying in a sales cycle. Maybe DTC cash pay at launch is part of that but only 1) insofar as it doesn’t distract you from the ultimate goal of proving value to payers/employers 2) you have the stomach for a long, hard road as a basically unproven business model (despite how great your topline might look).
Can there be exceptions to these frameworks? Certainly. Will there be exceptions? Maybe. The most likely area for success will be mental health as 1) it is the easiest to port over into an all virtual environment 2) it is a very high priority for payers and employers 3) people are used to paying for it out of pocket already 4) scarcity and access to care is actually a large problem on its own and 5) the bar to clear for 10x better software and products is extremely low. Brightside (a Slow company) is probably the best bet to make it work as a digital care provider.
¹ There are folks like ApolloMed and Privia in the “tech-enabled” healthcare camp, but they don’t grow organically nor do they cashflow super attractively at scale. They pretty much just look like plain old healthcare groups with nice(r) websites than most. It’s definitely not a bad model but it scales with people and M&A.
² There a number of one-time charges related to IPOs and/or mergers on one hand and pulled forward covid revenue on the other that might make these numbers look even worse BUT the trend goes back as far as 2017.
³ The one services-oriented approach we think could work is to lean into up/cross-sell built on proprietary infrastructure that can connect virtual to in-person care. Over time a company like Ro (a Slow port co) can increasingly become the connective tissue for digital health while developing a platform and a network with a single client: itself.
Many thanks to Daisy Wolf, Jomayra Herrera, Finn Murphy, Sharla Grass, Charlotte Ross, Kristin Baker Spohn, Everett Randle, Jake Swinghamer, Matt Gibstein, Jason Shuman, and Christian Garett for the feedback along the way.
Super cool to open up the draft to such a large group on the Slow Summer list.